Trained on CityScapes Dataset & Finetuned on KITTI Semantic dataset
# weight_path is default to checkpoints/BiseNetv2_150.pth
python3 demo.py --path PATH_TO_IMAGE --weight_path PATH_TO_BISENET_WEIGHTS
python3 visualization.py
python3 train.py --datapath PATH_TO_DATASET --pretrained PATH_TO_WEIGHTS_TO_CONTINUE --batch_size 16
python3 test.py --weight_path PATH_TO_WEIGHTS
# you must have KITTI dataset in data/KITTI in the format below
├── checkpoints
├── BiseNetv2_150.pth # path to model
├── tensorboard # path to save tensorboard events
├── data # path to kitti semantic dataset
├── KITTI
├── testing
├── image_2
├── training
├── image_2
├── instance
├── semantic
├── semantic_rgb