ReyaLabColumbia commited on
Commit
67be596
·
verified ·
1 Parent(s): 1d07707

Update README.md

Browse files

The Segformer-Cityscapes model was changed to a ternary classifier and fine-tuned on custom training data.
![example2.png](https://cdn-uploads.huggingface.co/production/uploads/6813e81cfd7a2af93d0e5384/-Igd3jirHou-ChM3ZHZvo.png)
![example.png](https://cdn-uploads.huggingface.co/production/uploads/6813e81cfd7a2af93d0e5384/N3b_rdSfWec4ybIB6Ijfq.png)

After training, it was able to correctly identify organoids and necrosis.
![test2.png](https://cdn-uploads.huggingface.co/production/uploads/6813e81cfd7a2af93d0e5384/UgaNJUKrRRwSxDTdXtuYD.png)
![Mask_test2.png](https://cdn-uploads.huggingface.co/production/uploads/6813e81cfd7a2af93d0e5384/Bt69LTm65lX76rzx2No19.png)

The python program (see linked GitHub) then uses the masks to annotate the images and provide statistics about the colonies.
![Group_analysis_results3.png](https://cdn-uploads.huggingface.co/production/uploads/6813e81cfd7a2af93d0e5384/dUcZcjmMp4jbRsSgQJLiD.png)

Files changed (1) hide show
  1. README.md +5 -0
README.md CHANGED
@@ -20,7 +20,12 @@ Custom fine-tuned version of NVIDIA's segformer model for colony slides in micro
20
  Xie, E., Wang, W., Yu, Z., Anandkumar, A., Alvarez, J. M., & Luo, P. (2021). SegFormer: Simple and efficient design for semantic segmentation with transformers. arXiv preprint arXiv:2105.15203. https://arxiv.org/abs/2105.15203
21
 
22
  The Segformer-Cityscapes model was changed to a ternary classifier and fine-tuned on custom training data.
 
 
23
 
24
  After training, it was able to correctly identify organoids and necrosis.
 
 
25
 
26
  The python program (see linked GitHub) then uses the masks to annotate the images and provide statistics about the colonies.
 
 
20
  Xie, E., Wang, W., Yu, Z., Anandkumar, A., Alvarez, J. M., & Luo, P. (2021). SegFormer: Simple and efficient design for semantic segmentation with transformers. arXiv preprint arXiv:2105.15203. https://arxiv.org/abs/2105.15203
21
 
22
  The Segformer-Cityscapes model was changed to a ternary classifier and fine-tuned on custom training data.
23
+ ![example2.png](https://cdn-uploads.huggingface.co/production/uploads/6813e81cfd7a2af93d0e5384/-Igd3jirHou-ChM3ZHZvo.png)
24
+ ![example.png](https://cdn-uploads.huggingface.co/production/uploads/6813e81cfd7a2af93d0e5384/N3b_rdSfWec4ybIB6Ijfq.png)
25
 
26
  After training, it was able to correctly identify organoids and necrosis.
27
+ ![test2.png](https://cdn-uploads.huggingface.co/production/uploads/6813e81cfd7a2af93d0e5384/UgaNJUKrRRwSxDTdXtuYD.png)
28
+ ![Mask_test2.png](https://cdn-uploads.huggingface.co/production/uploads/6813e81cfd7a2af93d0e5384/Bt69LTm65lX76rzx2No19.png)
29
 
30
  The python program (see linked GitHub) then uses the masks to annotate the images and provide statistics about the colonies.
31
+ ![Group_analysis_results3.png](https://cdn-uploads.huggingface.co/production/uploads/6813e81cfd7a2af93d0e5384/dUcZcjmMp4jbRsSgQJLiD.png)