We introduce Recurrent All-Pairs Field Transforms (RAFT), a new deep network architecture for estimating optical flow. RAFT extracts per-pixel features, builds multi-scale 4D correlation volumes for all pairs of pixels, and iteratively updates a flow field through a recurrent unit that performs lookups on the correlation volumes. RAFT achieves state-of-the-art performance on both KITTI and Sintel, with strong cross-dataset generalization and high efficiency in inference time, training speed, and parameter count.
Authors: Zachary Teed and Jia Deng @ Princeton University
Paper URL: https://arxiv.org/pdf/2003.12039.pdf
GitHub with code: https://github.com/princeton-vl/RAFT
Song credit: https://soundcloud.com/mattis-rodrigue/sans-titre