{
  "@context": "https://w3id.org/croissant/context.jsonld",
  "@type": "Dataset",
  "name": "MomentSeeker",
  "description": "A large-scale multi-modal benchmark for long video moment retrieval.",
  "license": "https://huggingface.co/datasets/avery00/MomentSeeker/blob/main/LICENSE",
  "creator": [
    {
      "name": "Huaying Yuan, Jian Ni"
    }
  ],
  "distribution": [
    {
      "contentUrl": "https://huggingface.co/datasets/avery00/MomentSeeker/resolve/main/videos.tar.gz.part_aa",
      "encodingFormat": "application/gzip"
    },
    {
      "contentUrl": "https://huggingface.co/datasets/avery00/MomentSeeker/resolve/main/videos.tar.gz.part_ab",
      "encodingFormat": "application/gzip"
    },
    {
      "contentUrl": "https://huggingface.co/datasets/avery00/MomentSeeker/resolve/main/videos.tar.gz.part_ac",
      "encodingFormat": "application/gzip"
    },
    {
      "contentUrl": "https://huggingface.co/datasets/avery00/MomentSeeker/resolve/main/query_images.tar.gz",
      "encodingFormat": "application/gzip"
    },
    {
      "contentUrl": "https://huggingface.co/datasets/avery00/MomentSeeker/resolve/main/query_videos.tar.gz",
      "encodingFormat": "application/gzip"
    },
    {
      "contentUrl": "https://huggingface.co/datasets/avery00/MomentSeeker/resolve/main/t2v.json",
      "encodingFormat": "application/json"
    },
    {
      "contentUrl": "https://huggingface.co/datasets/avery00/MomentSeeker/resolve/main/ti2v.json",
      "encodingFormat": "application/json"
    },
    {
      "contentUrl": "https://huggingface.co/datasets/avery00/MomentSeeker/resolve/main/tv2v.json",
      "encodingFormat": "application/json"
    }
  ],
  "field": [
    {
      "name": "qry_text",
      "description": "The textual query describing what to look for in the video.",
      "dataType": "string"
    },
    {
      "name": "qry_img_path",
      "description": "Path to the query image, if applicable.",
      "dataType": "string"
    },
    {
      "name": "qry_video_path",
      "description": "Path to the query video, if applicable.",
      "dataType": "string"
    },
    {
      "name": "src_video_path",
      "description": "Path to the source video in which to locate the answer.",
      "dataType": "string"
    },
    {
      "name": "task",
      "description": "Task type such as 'Description Location', 'Action Recognition', or 'Causal Reasoning'.",
      "dataType": "string"
    },
    {
      "name": "answering_time_interval",
      "description": "A list of [start_time, end_time] in seconds specifying the ground truth temporal segment.",
      "dataType": "array"
    }
  ],
  "data": [
    {
      "qry_text": "The video shows a man playing a game of Connect Four with a woman, who wins the game.",
      "qry_img_path": "",
      "qry_video_path": "",
      "src_video_path": "./videos/ego_79.mp4",
      "task": "Description Location",
      "answering_time_interval": [[415.0, 480.0]]
    },
    {
      "qry_text": "Did the man and the woman in the picture embrace?",
      "qry_img_path": "./query_images/movie101_77/frame_173.jpg",
      "qry_video_path": "",
      "src_video_path": "./videos/movie101_77.mp4",
      "task": "Action Recognition",
      "answering_time_interval": [[180, 198]]
    },
    {
      "qry_text": "What happened to this window afterward?",
      "qry_img_path": "",
      "qry_video_path": "./query_videos/tomjerry_8/177.00_181.00.mp4",
      "src_video_path": "./videos/tomjerry_8.mp4",
      "task": "Causal Reasoning",
      "answering_time_interval": [[181, 186]]
    }
  ]
}