SMURF: Self-Teaching Multi-Frame Unsupervised RAFT with Full-Image Warping

Austin Stone, Daniel Maurer, Alper Ayvaci, Anelia Angelova, Rico Jonschkowski

CVPR 2021 pp. 3887-3896

doi:10.1109/CVPR46437.2021.00388 /cvpr/2021/stone2021cvpr-smurf/

Abstract

We present SMURF, a method for unsupervised learning of optical flow that improves state of the art on all benchmarks by 36% to 40% and even outperforms several supervised approaches such as PWC-Net and FlowNet2. Our method integrates architecture improvements from supervised optical flow, i.e. the RAFT model, with new ideas for unsupervised learning that include a novel unsupervised sequence loss and self-supervision loss, a technique for handling out-of-frame motion, and an approach for learning effectively from multi-frame video data while still only requiring two frames for inference.

PDF CVPR Semantic Scholar

Cite

Text

Stone et al. "SMURF: Self-Teaching Multi-Frame Unsupervised RAFT with Full-Image Warping." Conference on Computer Vision and Pattern Recognition, 2021. doi:10.1109/CVPR46437.2021.00388

Markdown

[Stone et al. "SMURF: Self-Teaching Multi-Frame Unsupervised RAFT with Full-Image Warping." Conference on Computer Vision and Pattern Recognition, 2021.](https://mlanthology.org/cvpr/2021/stone2021cvpr-smurf/) doi:10.1109/CVPR46437.2021.00388

BibTeX

@inproceedings{stone2021cvpr-smurf,
  title     = {{SMURF: Self-Teaching Multi-Frame Unsupervised RAFT with Full-Image Warping}},
  author    = {Stone, Austin and Maurer, Daniel and Ayvaci, Alper and Angelova, Anelia and Jonschkowski, Rico},
  booktitle = {Conference on Computer Vision and Pattern Recognition},
  year      = {2021},
  pages     = {3887-3896},
  doi       = {10.1109/CVPR46437.2021.00388},
  url       = {https://mlanthology.org/cvpr/2021/stone2021cvpr-smurf/}
}