Temporally Distributed Networks for Fast Video Semantic Segmentation (CVPR'20)
Ping Hu, Fabian Caba Heilbron, Oliver Wang, Zhe Lin, Stan Sclaroff, Federico Perazzi
We present TDNet, a temporally distributed network designed for fast and accurate video semantic segmentation. We observe that features extracted from a certain high-level layer of a deep CNN can be approximated by composing features extracted from several shallower subnetworks. Leveraging the inherent temporal continuity in videos, we distribute these sub-networks over sequential frames. Therefore, at each time step, we only need to perform a lightweight computation to extract a sub-features group from a single sub-network. The full features used for segmentation are then recomposed by the application of a novel attention propagation module that compensates for geometry deformation between frames. A grouped knowledge distillation loss is also introduced to further improve the representation power at both full and sub-feature levels. Experiments on Cityscapes, CamVid, and NYUD-v2 demonstrate that our method achieves state-of-the-art accuracy with significantly faster speed and lower latency
- Linux
- Python 3.7
- Pytorch 1.1.0
- NVIDIA GPU + CUDA 10.0
pip install -r requirements.txtsee TEST_README.md
see TRAIN_README.md
If you find TDNet is helpful in your research, please consider citing:
@InProceedings{hu2020tdnet,
title={Temporally Distributed Networks for Fast Video Semantic Segmentation},
author={Hu, Ping and Caba, Fabian and Wang, Oliver and Lin, Zhe and Sclaroff, Stan and Perazzi, Federico},
journal={CVPR},
year={2020}
}
This is a pytorch re-implementation of TDNet, please refer to the original paper Temporally Distributed Networks for Fast Video Semantic Segmentation for comparisons.