Update README.md
Browse files
README.md
CHANGED
|
@@ -19,6 +19,8 @@ Custom fine-tuned version of NVIDIA's segformer model for colony slides in micro
|
|
| 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 |
|
|
|
|
|
|
|
| 22 |
The Segformer-Cityscapes model was changed to a ternary classifier and fine-tuned on custom training data, where colonies and necrosis were made as separate masks and then merged with different grayscale values.
|
| 23 |
|
| 24 |
This is our more advanced model that is trained on two additional types of organoid and should be more flexible with novel organoids (hence the abbreviation GP 'general purpose').
|
|
|
|
| 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 |
|
| 22 |
+
This model takes 512x512 images for segmentation.
|
| 23 |
+
|
| 24 |
The Segformer-Cityscapes model was changed to a ternary classifier and fine-tuned on custom training data, where colonies and necrosis were made as separate masks and then merged with different grayscale values.
|
| 25 |
|
| 26 |
This is our more advanced model that is trained on two additional types of organoid and should be more flexible with novel organoids (hence the abbreviation GP 'general purpose').
|