Chesebrough commited on
Commit
9c786ef
1 Parent(s): c9a8046

Update README.md

Browse files

Added results, example image input and output, and more complete code

Files changed (1) hide show
  1. README.md +54 -5
README.md CHANGED
@@ -16,16 +16,34 @@ widget:
16
 
17
  # DPT (large-sized model) fine-tuned on ADE20k
18
 
19
- Dense Prediction Transformer (DPT) model trained on ADE20k for semantic segmentation. It was introduced in the paper [Vision Transformers for Dense Prediction](https://arxiv.org/abs/2103.13413) by Ranftl et al. and first released in [this repository](https://github.com/isl-org/DPT).
20
 
21
- Disclaimer: The team releasing DPT did not write a model card for this model so this model card has been written by the Hugging Face team.
 
 
 
22
 
23
  ## Model description
24
 
25
- DPT uses the Vision Transformer (ViT) as backbone and adds a neck + head on top for semantic segmentation.
 
 
 
 
26
 
27
  ![model image](https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/dpt_architecture.jpg)
28
 
 
 
 
 
 
 
 
 
 
 
 
29
  ## Intended uses & limitations
30
 
31
  You can use the raw model for semantic segmentation. See the [model hub](https://huggingface.co/models?search=dpt) to look for
@@ -40,16 +58,47 @@ from transformers import DPTFeatureExtractor, DPTForSemanticSegmentation
40
  from PIL import Image
41
  import requests
42
 
43
- url = "http://images.cocodataset.org/val2017/000000039769.jpg"
44
  image = Image.open(requests.get(url, stream=True).raw)
45
 
46
- feature_extractor = DPTFeatureExtractor.from_pretrained("Intel/dpt-large-ade")
47
  model = DPTForSemanticSegmentation.from_pretrained("Intel/dpt-large-ade")
48
 
49
  inputs = feature_extractor(images=image, return_tensors="pt")
50
 
51
  outputs = model(**inputs)
52
  logits = outputs.logits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
53
  ```
54
 
55
  For more code examples, we refer to the [documentation](https://huggingface.co/docs/transformers/master/en/model_doc/dpt).
 
16
 
17
  # DPT (large-sized model) fine-tuned on ADE20k
18
 
19
+ The model is used for semantic segmentation of input images such as seen in the table below:
20
 
21
+
22
+ | Input Image | Output Segmented Image |
23
+ | --- | --- |
24
+ | ![input image](https://cdn-uploads.huggingface.co/production/uploads/641bd18baebaa27e0753f2c9/cG0alacJ4MeSL18CneD2u.png) | ![Segmented image](https://cdn-uploads.huggingface.co/production/uploads/641bd18baebaa27e0753f2c9/G3g6Bsuti60-bCYzgbt5o.png)|
25
 
26
  ## Model description
27
 
28
+ The Midas 3.0 nbased Dense Prediction Transformer (DPT) model was trained on ADE20k for semantic segmentation. It was introduced in the paper
29
+ [Vision Transformers for Dense Prediction](https://arxiv.org/abs/2103.13413) by Ranftl et al. and first released in [this repository](https://github.com/isl-org/DPT).
30
+
31
+
32
+ The MiDaS v3.0 DPT uses the Vision Transformer (ViT) as backbone and adds a neck + head on top for semantic segmentation.
33
 
34
  ![model image](https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/dpt_architecture.jpg)
35
 
36
+ Disclaimer: The team releasing DPT did not write a model card for this model so this model card has been written by the Hugging Face and the Intel AI Community team.
37
+
38
+
39
+ ## Results:
40
+ According to the authors, at the time of publication, when applied to semantic segmentation, dense vision transformers set a new state of the art on
41
+
42
+ **ADE20K with 49.02% mIoU.**
43
+
44
+ We further show that the architecture can be fine-tuned on smaller datasets such as NYUv2, KITTI, and Pascal Context where it also sets the new state of the art. Our models are available at
45
+ [Intel DPT GItHub Repository](https://github.com/intel-isl/DPT).
46
+
47
  ## Intended uses & limitations
48
 
49
  You can use the raw model for semantic segmentation. See the [model hub](https://huggingface.co/models?search=dpt) to look for
 
58
  from PIL import Image
59
  import requests
60
 
61
+ url = "http://images.cocodataset.org/val2017/000000026204.jpg"
62
  image = Image.open(requests.get(url, stream=True).raw)
63
 
64
+ feature_extractor = DPTImageProcessor .from_pretrained("Intel/dpt-large-ade")
65
  model = DPTForSemanticSegmentation.from_pretrained("Intel/dpt-large-ade")
66
 
67
  inputs = feature_extractor(images=image, return_tensors="pt")
68
 
69
  outputs = model(**inputs)
70
  logits = outputs.logits
71
+ print(logits.shape)
72
+ logits
73
+ prediction = torch.nn.functional.interpolate(
74
+ logits,
75
+ size=image.size[::-1], # Reverse the size of the original image (width, height)
76
+ mode="bicubic",
77
+ align_corners=False
78
+ )
79
+
80
+ # Convert logits to class predictions
81
+ prediction = torch.argmax(prediction, dim=1) + 1
82
+
83
+ # Squeeze the prediction tensor to remove dimensions
84
+ prediction = prediction.squeeze()
85
+
86
+ # Move the prediction tensor to the CPU and convert it to a numpy array
87
+ prediction = prediction.cpu().numpy()
88
+
89
+ # Convert the prediction array to an image
90
+ predicted_seg = Image.fromarray(prediction.squeeze().astype('uint8'))
91
+
92
+ # Define the ADE20K palette
93
+ adepallete = [0,0,0,120,120,120,180,120,120,6,230,230,80,50,50,4,200,3,120,120,80,140,140,140,204,5,255,230,230,230,4,250,7,224,5,255,235,255,7,150,5,61,120,120,70,8,255,51,255,6,82,143,255,140,204,255,4,255,51,7,204,70,3,0,102,200,61,230,250,255,6,51,11,102,255,255,7,71,255,9,224,9,7,230,220,220,220,255,9,92,112,9,255,8,255,214,7,255,224,255,184,6,10,255,71,255,41,10,7,255,255,224,255,8,102,8,255,255,61,6,255,194,7,255,122,8,0,255,20,255,8,41,255,5,153,6,51,255,235,12,255,160,150,20,0,163,255,140,140,140,250,10,15,20,255,0,31,255,0,255,31,0,255,224,0,153,255,0,0,0,255,255,71,0,0,235,255,0,173,255,31,0,255,11,200,200,255,82,0,0,255,245,0,61,255,0,255,112,0,255,133,255,0,0,255,163,0,255,102,0,194,255,0,0,143,255,51,255,0,0,82,255,0,255,41,0,255,173,10,0,255,173,255,0,0,255,153,255,92,0,255,0,255,255,0,245,255,0,102,255,173,0,255,0,20,255,184,184,0,31,255,0,255,61,0,71,255,255,0,204,0,255,194,0,255,82,0,10,255,0,112,255,51,0,255,0,194,255,0,122,255,0,255,163,255,153,0,0,255,10,255,112,0,143,255,0,82,0,255,163,255,0,255,235,0,8,184,170,133,0,255,0,255,92,184,0,255,255,0,31,0,184,255,0,214,255,255,0,112,92,255,0,0,224,255,112,224,255,70,184,160,163,0,255,153,0,255,71,255,0,255,0,163,255,204,0,255,0,143,0,255,235,133,255,0,255,0,235,245,0,255,255,0,122,255,245,0,10,190,212,214,255,0,0,204,255,20,0,255,255,255,0,0,153,255,0,41,255,0,255,204,41,0,255,41,255,0,173,0,255,0,245,255,71,0,255,122,0,255,0,255,184,0,92,255,184,255,0,0,133,255,255,214,0,25,194,194,102,255,0,92,0,255]
94
+
95
+ # Apply the color map to the predicted segmentation image
96
+ predicted_seg.putpalette(adepallete)
97
+
98
+ # Blend the original image and the predicted segmentation image
99
+ out = Image.blend(image, predicted_seg.convert("RGB"), alpha=0.5)
100
+
101
+ out
102
  ```
103
 
104
  For more code examples, we refer to the [documentation](https://huggingface.co/docs/transformers/master/en/model_doc/dpt).