kaggle link -> https://www.kaggle.com/code/banddaniel/iris-segmentation-u-net-w-tpu-dice-coef-0-94

I have used the following methods.

• Dice coefficient and Jaccard index implementation,
• The project took place using Google TPU,
• Custom layers for encoding and decoding,
• Custom callback class that used predicting a sample from the train dataset during training

• Test Dice Coefficient : 0.94
• Test Jaccard Index : 0.88

1. Ronneberger, O., Fischer, P., & Brox, T. (2015). U-Net: Convolutional Networks for Biomedical Image Segmentation (Version 1). arXiv. https://doi.org/10.48550/ARXIV.1505.04597
2. https://www.aao.org/eye-health/anatomy/parts-of-eye
3. https://en.wikipedia.org/wiki/Sørensen–Dice_coefficient