DHS net for salient objects detection(Pytorch implementation). Some part of this project is based on codes from https://github.com/wlguan/DHSNet-PyTorch , and this project is an optimized version.
Original running environment:
- Python 3.7.5
- Pytorch 1.3.1
- TorchVision 0.2.1
- pillow 7.0.0
See requirements.txt for detail.
- Put corresponding dataset in ./input/
- training images(RGB, jpg format): ./input/train/raw/
- training masks(gray, png format): ./input/train/mask/
- validation images(RGB, jpg format): ./input/test/raw/
- validation masks(gray, png format): ./input/test/mask/
- Run train.py, if you want to change some parameters, see train.py for detail.
- Put inference data in ./inference/
- inference images(RGB, jpg format): ./inference
- Run inference.py, output saliency maps will be in ./output directory.