UNet Corm & Damage Semantic Segmentation Model

This repository contains weights, performance curves, and hyperparameter logs for experiment unet_run_2026-06-26_18-26.

Model Summary

• Architecture Type: UNet (SMP wrapper)
• Backbone Encoder: resnet34
• Pretrained Weights: imagenet
• Input Channels: 3
• Number of Classes: 3 (0: background, 1: damage, 2: corm)
• Training Resolution: 512x512
• Loss Optimizer configuration: CrossEntropy

Metrics Curves

Below are the training performance plots generated for this run: