This repo is a PyTorch implementation of applying VAN (Visual Attention Network) to semantic segmentation. The code is based on mmsegmentaion.

More details can be found in Visual Attention Network.

@article{guo2022visual,
  title={Visual Attention Network},
  author={Guo, Meng-Hao and Lu, Cheng-Ze and Liu, Zheng-Ning and Cheng, Ming-Ming and Hu, Shi-Min},
  journal={arXiv preprint arXiv:2202.09741},
  year={2022}
}

Notes: Pre-trained models can be found in TsingHua Cloud.

Notes: In this scheme, we use multi-scale validation following Swin-Transformer. FLOPs are tested under the input size of 2048 $\times$ 512 using torchprofile (recommended, highly accurate and automatic MACs/FLOPs statistics).

Install MMSegmentation and download ADE20K according to the guidelines in MMSegmentation.

pip install mmsegmentation==0.26.0 (https://github.com/open-mmlab/mmsegmentation/tree/v0.26.0)

We use 8 GPUs for training by default. Run:

./dist_train.sh /path/to/config 8

To evaluate the model, run:

./dist_test.sh /path/to/config /path/to/checkpoint_file 8 --eval mIoU

Install torchprofile using

pip install torchprofile

To calculate FLOPs for a model, run:

bash tools/flops.sh /path/to/config --shape 512 512

Our implementation is mainly based on mmsegmentaion, Swin-Transformer, PoolFormer, and Enjoy-Hamburger. Thanks for their authors.

This repo is under the Apache-2.0 license. For commercial use, please contact the authors.