Maunil2k / UNet_research_paper_implementation

UNet: Convolutional Networks for Biomedical Image Segmentation Research paper implementation

UNet Research Paper Implementation

This repository shows a modified implementation of research paper: U-Net: Convolutional Networks for Biomedical Image Segmentation using Tensorflow. Further, the model was trained on HubMap HPA Dataset. The training results have been uploaded as well in the CSV file. Similarly, the loss chart and f1 score chart for the same are also uploaded.
Some modifications have been made in the implementation with respect to the model displayed in original paper. The size of the input image has been altered in order to get the same sized mask.

References:

1. Research paper: U-Net: Convolutional Networks for Biomedical Image Segmentation by Olaf Ronneberger, Philipp Fischer, and Thomas Brox.
2. Training Dataset: HuBMAP + HPA - Hacking the Human Body