Dataset Viewer
original_image
imagewidth (px) 512
512
| id
int64 4
4.26k
| image
imagewidth (px) 512
512
| conversations
listlengths 3
3
|
|---|---|---|---|
1,952 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What type of window treatment is featured on a plain, light-colored wall?",
"role": "user"
},
{
"content": "Light blue honeycomb blinds on windows with no other objects or textures in the background.",
"role": "assistant"
}
] |
||
430 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "Scene Context:\n- Objet : \"couch\", Position : in the middle on the right\n\n\nWhat style of furniture and decor would be commonly found in a coastal-themed room that incorporates elements of sailing and the sea?",
"role": "user"
},
{
"content": "A coastal living room with a nautical theme, a striped sofa, and a rattan chest as storage.",
"role": "assistant"
}
] |
||
2,829 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What is being featured on a window, where nothing else appears to be present?",
"role": "user"
},
{
"content": "Black wooden Venetian blinds on windows with no other objects or textures in the background.",
"role": "assistant"
}
] |
||
2,699 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "Scene Context:\n- Objet : \"couch\", Position : at the bottom on the left\n\n\nWhat is typically found in a warm and inviting home space?",
"role": "user"
},
{
"content": "A cozy living room with a fireplace, a plush rug, and a large mirror.",
"role": "assistant"
}
] |
||
634 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What are some common fixtures found in an elegant bathroom design?",
"role": "user"
},
{
"content": "A classic powder room with a pedestal sink, a modern matte black faucet, and a small framed painting.",
"role": "assistant"
}
] |
||
157 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What style of bathroom features a large soaking tub, minimalist design, natural stone flooring, and brings the outdoors in through a living plant?",
"role": "user"
},
{
"content": "A zen-inspired bathroom with a soaker tub, a low-profile toilet, and river rock tile flooring with a potted bamboo plant.",
"role": "assistant"
}
] |
||
2,645 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "Scene Context:\n- Objet : \"potted plant\", Position : in the middle on the right\n\n\nWhat furniture pieces are typically found in a modern living room?",
"role": "user"
},
{
"content": "A contemporary living room with a tufted Chesterfield sofa, a geometric coffee table, and a large plant in the corner.",
"role": "assistant"
}
] |
||
2,973 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "Scene Context:\n- Objet : \"bed\", Position : at the bottom in the center\n\n\nWhat is a typical arrangement of furniture found in a modern bedroom?",
"role": "user"
},
{
"content": "A contemporary bedroom with a platform bed, a sleek dresser, a gallery wall, and a floor mirror next to the dresser.",
"role": "assistant"
}
] |
||
2,780 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What is a common design element used to create the illusion of more space in a small hallway?",
"role": "user"
},
{
"content": "A minimalist hallway with overhead LED lighting, and a large round mirror.",
"role": "assistant"
}
] |
||
787 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "Scene Context:\n- Objet : \"couch\", Position : in the middle on the right\n\n\nWhat style of interior design is characterized by natural materials like wood and stone, often paired with comfortable seating areas?",
"role": "user"
},
{
"content": "A rustic living room with exposed wooden beams, a stone fireplace, and cozy upholstered armchairs.",
"role": "assistant"
}
] |
||
3,258 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What is the aesthetic of a room featuring vibrant textiles, antique flooring, and eclectic decorative elements?",
"role": "user"
},
{
"content": "A bohemian living room with colorful throw pillows and a vintage rug.",
"role": "assistant"
}
] |
||
2,722 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "Scene Context:\n- Objet : \"microwave\", Position : in the middle on the right\n- Objet : \"refrigerator\", Position : in the middle in the center\n- Objet : \"oven\", Position : in the middle on the right\n\n\nWhat are some common features found in modern kitchens that can contribute to a sense of luxury and functionality?",
"role": "user"
},
{
"content": "A spacious kitchen with marble countertops, stainless steel appliances, and an undermount sink.",
"role": "assistant"
}
] |
||
1,111 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What style of furniture does this bright living room feature?",
"role": "user"
},
{
"content": "A bright living room with white walls, a navy blue sofa, a marble-topped coffee table, and a rug with geometric patterns.",
"role": "assistant"
}
] |
||
1,798 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "Scene Context:\n- Objet : \"couch\", Position : in the middle on the left\n\n\nWhat is typically found in a warm and inviting living space featuring a neutral-colored seating area, wooden furniture, and plenty of literature and plants?",
"role": "user"
},
{
"content": "A cozy living room with a beige loveseat, a rustic coffee table, and a bookshelf full of books with a green plant on top.",
"role": "assistant"
}
] |
||
261 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What color is an oak wood coffee table typically placed against?",
"role": "user"
},
{
"content": "An oak wood coffee table in a white background with no other objects or textures.",
"role": "assistant"
}
] |
||
3,021 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "Scene Context:\n- Objet : \"potted plant\", Position : at the bottom on the right\n\n\nWhat is a common setup found on a cozy outdoor space?",
"role": "user"
},
{
"content": "A rustic porch with a swing bench, potted plants, string lights, and a vintage chandelier.",
"role": "assistant"
}
] |
||
3,923 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "Scene Context:\n- Objet : \"chair\", Position : at the bottom in the center\n\n\nWhat furniture is typically found in a quiet, academic space designed for reading or studying?",
"role": "user"
},
{
"content": "A cozy study room with a large bookshelf, a comfortable armchair, and a contemporary desk.",
"role": "assistant"
}
] |
||
3,260 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What is the primary visual element being described in this image?",
"role": "user"
},
{
"content": "Charcoal gray Roman shades on windows with no other objects or textures in the background.",
"role": "assistant"
}
] |
||
3,518 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What style of furniture is typically characterized by its simple, clean lines and lack of distracting background elements?",
"role": "user"
},
{
"content": "A contemporary bar table with a wooden frame in a white background with no other objects or textures in the background.",
"role": "assistant"
}
] |
||
1,612 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What type of furniture is characterized by its simple, minimalist design featuring a plain white surface and ornate wooden legs?",
"role": "user"
},
{
"content": "A classic wooden console table with ornate legs in a white background with no other objects or textures.",
"role": "assistant"
}
] |
||
138 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What is the aesthetic style of a living room characterized by simplicity, a neutral color palette, and a mix of abstract art?",
"role": "user"
},
{
"content": "A minimalist living room with a low-profile couch, a fluffy white rug, and two black and white paintings on the wall.",
"role": "assistant"
}
] |
||
672 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "Scene Context:\n- Objet : \"refrigerator\", Position : in the middle on the right\n- Objet : \"microwave\", Position : in the middle on the left\n\n\nWhat features would make a typical modern kitchen feel luxurious?",
"role": "user"
},
{
"content": "A spacious kitchen with marble countertops, stainless steel appliances, and a large island.",
"role": "assistant"
}
] |
||
2,650 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What style of modern living space typically features large windows, a minimalist seating area, and a transparent surface for holding objects?",
"role": "user"
},
{
"content": "A contemporary loft living room with floor-to-ceiling windows, a sleek black sectional, and a glass coffee table.",
"role": "assistant"
}
] |
||
3,251 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What type of decorative elements would typically be used to add visual interest to an otherwise plain, monochromatic dining table setup featuring a concrete top and white background?",
"role": "user"
},
{
"content": "An industrial-style dining table with a concrete top in a white background without any other objects or textures.",
"role": "assistant"
}
] |
||
202 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What color and style of window treatments would be most visible when placed against a plain, featureless background?",
"role": "user"
},
{
"content": "Beige roller blinds on windows with no other objects or textures in the background.",
"role": "assistant"
}
] |
||
2,105 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "Scene Context:\n- Objet : \"tv\", Position : at the top on the left\n\n\nWhat is typically found on the floor in a relaxing living space?",
"role": "user"
},
{
"content": "A cozy living room with a comfortable sofa and a plush area rug.",
"role": "assistant"
}
] |
||
3,002 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What type of fixtures and materials are commonly used to create a traditional and cozy bathroom design?",
"role": "user"
},
{
"content": "A rustic bathroom with a stone accent wall, a wooden vanity, and a freestanding cast-iron bathtub.",
"role": "assistant"
}
] |
||
4,123 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What style of kitchen is characterized by its use of natural materials like wood, stone, and metal?",
"role": "user"
},
{
"content": "A rustic kitchen with natural wood elements, stone accents, brushed nickel hardware, and a farmhouse sink.",
"role": "assistant"
}
] |
||
46 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What style of decor is typically characterized by the use of floor cushions, lush greenery, and handmade textiles?",
"role": "user"
},
{
"content": "A bohemian living room with floor pillows, hanging plants, and a macrame wall hanging.",
"role": "assistant"
}
] |
||
79 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "Scene Context:\n- Objet : \"bench\", Position : at the bottom on the right\n\n\nWhat design elements typically accompany modern hallways to create a sense of calm and functionality?",
"role": "user"
},
{
"content": "A contemporary hallway with recessed lighting, and a simple bench with storage underneath.",
"role": "assistant"
}
] |
||
3,513 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What would be a key design element in this modern living room?",
"role": "user"
},
{
"content": "A contemporary living room with a statement art piece, a leather recliner, and a glass coffee table.",
"role": "assistant"
}
] |
||
3,984 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "Scene Context:\n- Objet : \"couch\", Position : at the bottom in the center\n- Objet : \"potted plant\", Position : at the bottom on the left\n\n\nWhat style would you expect to find in this eclectic yet harmonious living room setup?",
"role": "user"
},
{
"content": "A vintage styled living room with wooden floor, a Victorian sofa, and a modern art piece on the wall.",
"role": "assistant"
}
] |
||
3,014 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What are the typical fixtures commonly found in traditional bathrooms?",
"role": "user"
},
{
"content": "A classic bathroom with a clawfoot tub, pedestal sink, and a small window without curtains.",
"role": "assistant"
}
] |
||
2,835 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What are some essential elements typically found in a minimalist home living space?",
"role": "user"
},
{
"content": "A minimalist living room with a single chaise lounge, a hanging swing chair in the corner, a geometric area rug, and white walls.",
"role": "assistant"
}
] |
||
1,396 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "Scene Context:\n- Objet : \"chair\", Position : at the bottom in the center\n- Objet : \"potted plant\", Position : at the bottom on the left\n\n\nWhat is the typical aesthetic of a sparse, low-profile living space featuring a sleek armchair, a compact side table, and delicate window coverings?",
"role": "user"
},
{
"content": "A minimalist living room with a black Barcelona chair, a small round coffee table, and sheer curtains.",
"role": "assistant"
}
] |
||
1,749 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What is the design aesthetic of this bedroom characterized by simplicity, natural light, and earthy tones?",
"role": "user"
},
{
"content": "A minimalist bedroom with a platform bed, floor-to-ceiling windows, a neutral color palette, and a hanging plant for a touch of greenery.",
"role": "assistant"
}
] |
||
2,287 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What design elements characterize this compact bathroom, which incorporates both functionality and aesthetic appeal?",
"role": "user"
},
{
"content": "A compact bathroom featuring a corner shower with a frosted glass door, a small sink, and a nautical-themed painting on the wall above the toilet.",
"role": "assistant"
}
] |
||
4,215 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What are some key elements that might be found in a lively children's play area?",
"role": "user"
},
{
"content": "A vibrant playroom with colorful toy storage, a mini basketball hoop, and a reading nook.",
"role": "assistant"
}
] |
||
1,230 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What type of furniture is typically found on a bedside table and what style of handle does it usually feature?",
"role": "user"
},
{
"content": "A classic bedside table with simple, modern handles in a white background.",
"role": "assistant"
}
] |
||
3,094 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What is typically found in a modern bathroom that features white tile walls and a large reflective surface?",
"role": "user"
},
{
"content": "A bright bathroom with white subway tiles, a glass-door shower, a large vanity mirror, and a single sink with a green potted plant on the vanity.",
"role": "assistant"
}
] |
||
3,593 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What features are typically found in a high-end outdoor living space designed for relaxation and ambiance?",
"role": "user"
},
{
"content": "A luxury patio area with outdoor seating, a fire pit, a pergola with string lights, and outdoor lighting.",
"role": "assistant"
}
] |
||
4,198 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What style or aesthetic is typically associated with a traditional living room featuring natural materials like hardwood floors, leather upholstery, and stone fireplaces?",
"role": "user"
},
{
"content": "A rustic living room with hardwood floors, leather sofas, and a stone fireplace with a vintage clock as mantel decor.",
"role": "assistant"
}
] |
||
4,189 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What is a common design aesthetic for a kitchen characterized by a predominantly white color scheme, incorporated natural elements, and an emphasis on bringing in plants?",
"role": "user"
},
{
"content": "Scandinavian-inspired kitchen with white walls, wood accents, natural light, and pops of greenery.",
"role": "assistant"
}
] |
||
731 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "Scene Context:\n- Objet : \"couch\", Position : at the bottom in the center\n\n\nWhat style would you expect to find in a vintage-inspired home interior featuring a floral patterned wall decoration?",
"role": "user"
},
{
"content": "A traditional living room with floral wallpaper, antique furniture, and a classic chandelier.",
"role": "assistant"
}
] |
||
512 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What type of fixtures are typically found together in an antique bathroom?",
"role": "user"
},
{
"content": "A vintage bathroom with a clawfoot bathtub and a pedestal sink.",
"role": "assistant"
}
] |
||
3,052 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "Scene Context:\n- Objet : \"couch\", Position : at the bottom on the left\n\n\nWhat style of home decor features a navy-blue sofa as a central piece, often paired with a fireplace and decorative artwork?",
"role": "user"
},
{
"content": "A cozy living room with a navy-blue sofa facing a fireplace, framed artwork on the walls, and wooden side tables.",
"role": "assistant"
}
] |
||
3,173 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "Scene Context:\n- Objet : \"couch\", Position : in the middle on the left\n\n\nWhat design elements are typically found in a lively and inviting living room?",
"role": "user"
},
{
"content": "A vibrant living room with a multicolored area rug, a charcoal gray sofa, and bright yellow walls.",
"role": "assistant"
}
] |
||
2,072 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "Scene Context:\n- Objet : \"potted plant\", Position : at the bottom in the center\n\n\nWhat color and style of window treatment would be most visible against a completely plain backdrop?",
"role": "user"
},
{
"content": "Light gray zebra blinds on windows with no other objects or textures in the background.",
"role": "assistant"
}
] |
||
288 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What type of window treatment features a clean, minimalist aesthetic without any distracting visual elements?",
"role": "user"
},
{
"content": "White painted wood shades on windows with no other objects or textures in the background.",
"role": "assistant"
}
] |
||
786 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What is a typical design element of a contemporary industrial bathroom?",
"role": "user"
},
{
"content": "A modern industrial bathroom with a concrete countertop, a copper vessel sink, and an exposed brick accent wall.",
"role": "assistant"
}
] |
||
1,717 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "Scene Context:\n- Objet : \"potted plant\", Position : at the bottom on the left\n\n\nWhat are some key elements typically found in a peaceful and inviting reading space?",
"role": "user"
},
{
"content": "A cozy reading corner with a comfy armchair, a pendant light for extra lighting, and a small side table.",
"role": "assistant"
}
] |
||
551 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What are some common features found in luxurious bathrooms?",
"role": "user"
},
{
"content": "A serene master bathroom with a soaking tub, a waterfall shower, and marble countertops.",
"role": "assistant"
}
] |
||
1,697 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What is the typical aesthetic of a traditional country home?",
"role": "user"
},
{
"content": "A rustic living room with exposed wooden beams, a fabric armchair, and a cozy fireplace.",
"role": "assistant"
}
] |
||
1,650 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "Scene Context:\n- Objet : \"couch\", Position : at the bottom in the center\n\n\nWhat type of window treatment is most visible when there are no other visual elements present?",
"role": "user"
},
{
"content": "Solid color roman blinds on windows with no other objects or texture in the background.",
"role": "assistant"
}
] |
||
2,601 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "Scene Context:\n- Objet : \"microwave\", Position : at the top in the center\n\n\nWhat style of kitchen features stainless steel appliances, granite countertops, and minimalist cabinetry?",
"role": "user"
},
{
"content": "A modern kitchen with stainless steel appliances, granite countertops, and sleek white cabinets.",
"role": "assistant"
}
] |
||
1,910 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What is typically found in a relaxing living room setup?",
"role": "user"
},
{
"content": "A cozy living room with a plush rug, a rocking chair, and a bookshelf filled with books.",
"role": "assistant"
}
] |
||
1,381 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "Scene Context:\n- Objet : \"chair\", Position : at the bottom in the center\n\n\nWhat are some key features typically found in a modern kitchen that serves both counter space and seating?",
"role": "user"
},
{
"content": "A spacious kitchen with an island, bar stools, and dark wood cabinets.",
"role": "assistant"
}
] |
||
3,731 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What type of material and color combination is typically used for this piece of furniture?",
"role": "user"
},
{
"content": "A marble coffee table with white background.",
"role": "assistant"
}
] |
||
3,336 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What is the typical aesthetic of a coastal-themed living room?",
"role": "user"
},
{
"content": "A coastal living room with a blue color scheme, a white linen sofa, and seashell decorations.",
"role": "assistant"
}
] |
||
1,393 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What style of decor would feature beige furniture, shells, and maritime-inspired elements to evoke a coastal atmosphere?",
"role": "user"
},
{
"content": "A coastal living room with neutral beige furniture, seashell decorations, and a nautical theme.",
"role": "assistant"
}
] |
||
3,123 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What is being depicted by a pair of solid color curtains hung from windows, devoid of any surrounding objects or visual interest?",
"role": "user"
},
{
"content": "Solid color tie-up shade curtains on windows with no other objects or textures in the background.",
"role": "assistant"
}
] |
||
4,182 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What is the appearance of a small, circular dining area set against a neutral-colored wall?",
"role": "user"
},
{
"content": "A round bistro table with a wooden base in a white background.",
"role": "assistant"
}
] |
||
1,705 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What is the visual contrast between a traditional Victorian-style piece of furniture and a minimalist modern object, presented against a neutral white background?",
"role": "user"
},
{
"content": "A Victorian-style end table with a simple modern design in a white background with no other objects or textures.",
"role": "assistant"
}
] |
||
1,387 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What is typically found in a stylishly decorated living room featuring a prominent seating area, bold color accents, and modern art?",
"role": "user"
},
{
"content": "A chic living room with white marble flooring, an emerald green velvet armchair, and a large abstract canvas artwork.",
"role": "assistant"
}
] |
||
1,302 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What features would be present in a modern living room that prioritizes comfort and aesthetics over traditional cooling methods?",
"role": "user"
},
{
"content": "An airy living room with an open-concept design, a sleek coffee table, without a wall-mounted air conditioning unit.",
"role": "assistant"
}
] |
||
1,195 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "Scene Context:\n- Objet : \"potted plant\", Position : in the middle on the left\n- Objet : \"couch\", Position : at the bottom on the right\n- Objet : \"couch\", Position : in the middle on the left\n\n\nWhat style of interior design is characterized by an emphasis on individuality, eclecticism, and a mix of vintage and global elements?",
"role": "user"
},
{
"content": "A bohemian living room with eclectic furniture, vibrant colors, and a hammock.",
"role": "assistant"
}
] |
||
1,077 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What is commonly found in a decorative room with soft colors and unique storage?",
"role": "user"
},
{
"content": "A narrow passageway with pastel-colored paint, a floating shelf displaying souvenirs, and a small ottoman.",
"role": "assistant"
}
] |
||
685 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What type of window treatment is being showcased in an image where the focus is solely on the blind, without any distracting elements?",
"role": "user"
},
{
"content": "Vertical fabric blinds on windows with no other objects or textures in the background.",
"role": "assistant"
}
] |
||
3,135 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What style would be characterized by a modern-looking living room featuring high-gloss furniture, a flat-screen television mounted on the wall, and hardwood flooring?",
"role": "user"
},
{
"content": "A sleek living room with high-gloss furniture, a mounted TV, and a bamboo floor mat.",
"role": "assistant"
}
] |
||
4,216 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What is typically found in a quiet space designed for relaxation and reading?",
"role": "user"
},
{
"content": "A cozy reading nook with a rocking chair, a floor lamp, and a bookshelf.",
"role": "assistant"
}
] |
||
946 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What would be described as being made entirely of a single color, in a plain setting, without any surrounding elements?",
"role": "user"
},
{
"content": "An outdoor patio table made of stainless steel in a white background with no other objects or textures.",
"role": "assistant"
}
] |
||
393 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What can be observed visually when a brightly colored curtain is hung between two empty, plain surfaces?",
"role": "user"
},
{
"content": "Brightly colored curtains on windows with no other objects or textures in the background.",
"role": "assistant"
}
] |
||
3,976 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What style of furniture and decor would be typical in a coastal-inspired home?",
"role": "user"
},
{
"content": "A beach-themed living room with a white slip-covered sofa, driftwood accents, and a ceiling fan.",
"role": "assistant"
}
] |
||
1,539 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What style aesthetic is typically associated with a room that features a plush velvet seating area, unconventional art pieces, and ornate metallic lighting fixtures?",
"role": "user"
},
{
"content": "A chic living room with a velvet loveseat, an abstract painting, and a brass lamp with a black lamp shade.",
"role": "assistant"
}
] |
||
3,788 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What style of decor would typically feature a plush velvet sofa, an original artwork, and a stylish lamp without a glass shade?",
"role": "user"
},
{
"content": "A chic living room with a velvet loveseat, an abstract painting, and a brass lamp without a shade.",
"role": "assistant"
}
] |
||
3,748 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "Scene Context:\n- Objet : \"couch\", Position : at the bottom in the center\n\n\nWhat style is characterized by a focus on clean lines, neutral colors, and limited decorative elements, as exemplified by this living room?",
"role": "user"
},
{
"content": "A minimalist living room with a simple leather couch, a floor lamp, and a geometric-patterned area rug.",
"role": "assistant"
}
] |
||
3,252 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What are some key elements typically found in a small-scale, stylish living room design?",
"role": "user"
},
{
"content": "A minimalist living room with a low-profile sofa, a glass coffee table, and accent chairs.",
"role": "assistant"
}
] |
||
1,198 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What is a typical setup for a contemporary living room that features a statement piece of furniture and decorative elements?",
"role": "user"
},
{
"content": "A modern living room with a large sectional sofa, a glass coffee table, and vibrant artwork.",
"role": "assistant"
}
] |
||
774 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What style of interior design is characterized by clean lines, minimal ornamentation, and an emphasis on functionality?",
"role": "user"
},
{
"content": "A chic foyer with a geometric patterned wallpaper, a mid-century modern console table, and a modern floor lamp.",
"role": "assistant"
}
] |
||
475 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What can be found in a typical luxurious bathroom that also features a relaxing soaking area?",
"role": "user"
},
{
"content": "A serene master bathroom with a soaking tub, a double vanity, and a large window.",
"role": "assistant"
}
] |
||
2,705 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "Scene Context:\n- Objet : \"couch\", Position : at the bottom in the center\n\n\nWhat is the typical arrangement of furniture and decor in a small to medium-sized living room?",
"role": "user"
},
{
"content": "A cozy living room with a beige sofa, a wooden coffee table, without a bookshelf.",
"role": "assistant"
}
] |
||
1,750 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "Scene Context:\n- Objet : \"couch\", Position : at the bottom in the center\n\n\nWhat style of interior design is characterized by the use of industrial elements such as exposed brick, metal accents, and leather upholstery in a living room?",
"role": "user"
},
{
"content": "An industrial living room with exposed brick walls, a cowhide rug, a metal coffee table, and leather sofas.",
"role": "assistant"
}
] |
||
3,148 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "Scene Context:\n- Objet : \"refrigerator\", Position : in the middle on the right\n\n\nWhat features would make up a modern and well-equipped kitchen?",
"role": "user"
},
{
"content": "A spacious kitchen with stainless steel appliances, marble countertops, and a kitchen island with bar stools.",
"role": "assistant"
}
] |
||
1,839 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What design elements are present in a simple and modern living space that incorporates vibrant colors and clever compartmentalization?",
"role": "user"
},
{
"content": "A minimalist living room with clean lines, pops of bright yellow, and hidden storage.",
"role": "assistant"
}
] |
||
1,689 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What style of furniture and decor is typically found in a space that features an industrial aesthetic?",
"role": "user"
},
{
"content": "An industrial-style living room with a metal coffee table, exposed brick walls, and a beige fabric sectional.",
"role": "assistant"
}
] |
||
1,465 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "Scene Context:\n- Objet : \"tv\", Position : at the top in the center\n\n\nWhat design elements are typically found in a modern living room?",
"role": "user"
},
{
"content": "A contemporary living room with a floating TV stand, a fabric sectional sofa, and a statement wall art.",
"role": "assistant"
}
] |
||
3,022 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What style is the living room decor inspired by, featuring a tufted leather sofa, Persian rug, and bookshelf with an antique clock?",
"role": "user"
},
{
"content": "A vintage-inspired living room with a tufted leather sofa, a Persian rug, and a bookshelf with an antique clock on it.",
"role": "assistant"
}
] |
||
4,034 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "Scene Context:\n- Objet : \"couch\", Position : at the bottom on the right\n- Objet : \"potted plant\", Position : in the middle on the left\n- Objet : \"couch\", Position : at the bottom on the left\n- Objet : \"tv\", Position : in the middle in the center\n\n\nWhat style aesthetic is represented by a modern living room featuring a unique lighting fixture, minimalist television display unit, and thick, textured floor covering?",
"role": "user"
},
{
"content": "A contemporary living room with a cluster pendant light, a sleek TV stand, and a shaggy rug.",
"role": "assistant"
}
] |
||
3,430 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What is being shaded by these blinds?",
"role": "user"
},
{
"content": "Bamboo blinds on windows with no other objects or textures in the background.",
"role": "assistant"
}
] |
||
1,682 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What are the essential features typically found in a modern bathroom that prioritize simplicity and elegance?",
"role": "user"
},
{
"content": "A minimalist bathroom with a single vanity, a frameless mirror, a walk-in shower, and a small potted plant on the vanity.",
"role": "assistant"
}
] |
||
1,212 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "Scene Context:\n- Objet : \"couch\", Position : in the middle in the center\n\n\nWhat is the typical design element often used to enhance natural light in a modern living room?",
"role": "user"
},
{
"content": "A modern living room with floor-to-ceiling windows, a leather armchair, a glass coffee table, and a colorful accent rug beneath.",
"role": "assistant"
}
] |
||
2,516 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What are some essential elements for creating a relaxing and intimate reading space?",
"role": "user"
},
{
"content": "A cozy reading nook with a hanging pendant light, an armchair, and a small bookshelf filled with books.",
"role": "assistant"
}
] |
||
867 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What is the typical architectural feature of a wealthy or aristocratic home characterized by a high ceiling, a sparkling light fixture, and a staircase that doesn't have a traditional runner?",
"role": "user"
},
{
"content": "A luxurious foyer with high ceilings, an ornate crystal chandelier, and a grand staircase without a runner.",
"role": "assistant"
}
] |
||
3,588 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "Scene Context:\n- Objet : \"couch\", Position : in the middle on the right\n- Objet : \"potted plant\", Position : in the middle on the left\n\n\nWhat is the typical furniture arrangement for a contemporary living room?",
"role": "user"
},
{
"content": "A modern living room with a sleek black TV stand, a navy blue sectional sofa, and a glass coffee table.",
"role": "assistant"
}
] |
||
3,875 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "Scene Context:\n- Objet : \"couch\", Position : at the top on the right\n\n\nWhat is the visual style of this empty room featuring a simple wooden coffee table on a glass surface?",
"role": "user"
},
{
"content": "A rustic wooden coffee table with a glass top in a white background with no other objects or textures in the background.",
"role": "assistant"
}
] |
||
3,184 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "Scene Context:\n- Objet : \"chair\", Position : at the bottom on the left\n\n\nWhat kind of furniture is typically found in a small, uncluttered dining space?",
"role": "user"
},
{
"content": "A minimalist dining room with a round table and simple chairs.",
"role": "assistant"
}
] |
||
54 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What is typically found in a traditionally decorated home that creates a warm ambiance?",
"role": "user"
},
{
"content": "A cozy living room with a fireplace, a plush rug, and a large mirror.",
"role": "assistant"
}
] |
||
928 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What type of aesthetic is typically characterized by vintage furniture, musical instruments from an earlier era, and classic photographic prints?",
"role": "user"
},
{
"content": "A vintage-inspired living room with retro furniture, a record player, and framed black and white photographs.",
"role": "assistant"
}
] |
||
1,237 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What is typically found in a luxurious bathroom designed for relaxation?",
"role": "user"
},
{
"content": "A serene bathroom with a deep soaking tub, a rainfall showerhead, a vanity with a mirror, and a potted plant on the windowsill.",
"role": "assistant"
}
] |
||
83 |
[
{
"content": "You are a vision-language assistant. Your task is to answer questions based on the visual context provided, addressed to a visually impaired person.\n Each scene is presented through a sequence of images, and each image is labeled with a number (e.g., \"Image 0\", \"Image 1\", etc.).\n It's very important to know that the images are ordered chronologically: \"Image 0\" is the first image, and the last image (Image 1) represents the current position of the person.\n\n Each image includes:\n - The object label (what the object is)\n - The approximate position in the image (top/middle/bottom and left/center/right)\n\n Use both the visual content of the images and the spatial information provided to understand the environment and guide with spatial instructions the person appropriately.\n Do not use visual information to guide but number of steps etc instead.\n\n Here is the scene context and question:\n ",
"role": "system"
},
{
"content": "What design elements are present in this simple and sleek living space?",
"role": "user"
},
{
"content": "A minimalist living room with a low-profile media console, a black area rug, and floor-to-ceiling windows.",
"role": "assistant"
}
] |
End of preview. Expand
in Data Studio
README.md exists but content is empty.
- Downloads last month
- 146