TorchSeg -- Fast, modular reference implementation and easy training of Semantic Segmentation algorithms in PyTorch

Hi all,
We have released the TorchSeg.
In this repo, we will release the fast, modular reference implementation of semantic segmentation algorithms in PyTorch.
Now, this repo has implemented some state-of-art semantic segmentation algorithms: PSPNet, DFN, BiSeNet, PSANet.
More algorithms will be supported.


  • Modular Design: easily construct customized semantic segmentation models by combining different components.
  • Distributed Training: >60% faster than the multi-thread parallel method(nn.DataParallel), we use the multi-processing parallel method.
  • Multi-GPU training and inference: support different manners of inference.
  • Provides pre-trained models and implement different semantic segmentation models.