Using a U-net for semantic segmentation
Dataset
The dataset should have the following directories
|train_set
|----|--original_images
|----|--masks
|test_set
|----|--original_images
|----|--masks
|val_set
|----|--original_images
|----|--masks
Validation Image
Random Google Images
Live Camera Feed Run the live_segmentation.py script for live segmentation after checking for where the video device is connected. Unix systems have video streming devices like webcams connected at /dev/video...
