nielsr HF staff commited on
Commit
dc65c96
1 Parent(s): 18819d6

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +7 -8
README.md CHANGED
@@ -34,7 +34,7 @@ You can use the raw model for object detection. See the [model hub](https://hugg
34
  Here is how to use this model:
35
 
36
  ```python
37
- from transformers import AutoFeatureExtractor, ConditionalDetrForObjectDetection
38
  import torch
39
  from PIL import Image
40
  import requests
@@ -42,24 +42,23 @@ import requests
42
  url = "http://images.cocodataset.org/val2017/000000039769.jpg"
43
  image = Image.open(requests.get(url, stream=True).raw)
44
 
45
- feature_extractor = AutoFeatureExtractor.from_pretrained("microsoft/conditional-detr-resnet-50")
46
  model = ConditionalDetrForObjectDetection.from_pretrained("microsoft/conditional-detr-resnet-50")
47
 
48
- inputs = feature_extractor(images=image, return_tensors="pt")
49
  outputs = model(**inputs)
50
 
51
  # convert outputs (bounding boxes and class logits) to COCO API
 
52
  target_sizes = torch.tensor([image.size[::-1]])
53
- results = feature_extractor.post_process(outputs, target_sizes=target_sizes)[0]
54
 
55
  for score, label, box in zip(results["scores"], results["labels"], results["boxes"]):
56
  box = [round(i, 2) for i in box.tolist()]
57
- # let's only keep detections with score > 0.7
58
- if score > 0.7:
59
- print(
60
  f"Detected {model.config.id2label[label.item()]} with confidence "
61
  f"{round(score.item(), 3)} at location {box}"
62
- )
63
  ```
64
  This should output:
65
  ```
 
34
  Here is how to use this model:
35
 
36
  ```python
37
+ from transformers import AutoImageProcessor, ConditionalDetrForObjectDetection
38
  import torch
39
  from PIL import Image
40
  import requests
 
42
  url = "http://images.cocodataset.org/val2017/000000039769.jpg"
43
  image = Image.open(requests.get(url, stream=True).raw)
44
 
45
+ processor = AutoImageProcessor.from_pretrained("microsoft/conditional-detr-resnet-50")
46
  model = ConditionalDetrForObjectDetection.from_pretrained("microsoft/conditional-detr-resnet-50")
47
 
48
+ inputs = processor(images=image, return_tensors="pt")
49
  outputs = model(**inputs)
50
 
51
  # convert outputs (bounding boxes and class logits) to COCO API
52
+ # let's only keep detections with score > 0.7
53
  target_sizes = torch.tensor([image.size[::-1]])
54
+ results = processor.post_process_object_detection(outputs, target_sizes=target_sizes, threshold=0.7)[0]
55
 
56
  for score, label, box in zip(results["scores"], results["labels"], results["boxes"]):
57
  box = [round(i, 2) for i in box.tolist()]
58
+ print(
 
 
59
  f"Detected {model.config.id2label[label.item()]} with confidence "
60
  f"{round(score.item(), 3)} at location {box}"
61
+ )
62
  ```
63
  This should output:
64
  ```