Update README.md
Browse filesThe SegFormer Cityscapes model was changed to a ternary classifier (source script in GitHub) and fine-tuned on a custom made dataset.


The trained model was then able to correctly identify colonies and necrosis from microscope images.


The python program (see linked GitHub) then uses these masks to circle colonies, necrosis, and provide statistics.

@@ -17,4 +17,5 @@ https://reya-lab.org/
|
|
17 |
|
18 |
Custom fine-tuned version of NVIDIA's segformer model for colony slides in microscopy.
|
19 |
|
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
|
|
|
|
17 |
|
18 |
Custom fine-tuned version of NVIDIA's segformer model for colony slides in microscopy.
|
19 |
|
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 |
+
|