Skip to main content

RAFT: Recurrent All-Pairs Field Transforms for Optical Flow

Author(s): Teed, Zachary; Deng, Jia

To refer to this page use:
Abstract: We introduce Recurrent All-Pairs Field Transforms (RAFT), a new deep network architecture for 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 KITTI, RAFT achieves an F1-all error of 5.10%, a 16% error reduction from the best published result (6.10%). On Sintel (final pass), RAFT obtains an end-point-error of 2.855 pixels, a 30% error reduction from the best published result (4.098 pixels). In addition, RAFT has strong cross-dataset generalization as well as high efficiency in inference time, training speed, and parameter count. Code is available at
Publication Date: 2020
Citation: Teed, Zachary, and Jia Deng. "RAFT: Recurrent All-Pairs Field Transforms for Optical Flow." European Conference on Computer Vision (ECCV) (2020): pp. 402-419. doi:10.1007/978-3-030-58536-5_24
DOI: 10.1007/978-3-030-58536-5_24
ISSN: 0302-9743
EISSN: 1611-3349
Pages: 402 - 419
Type of Material: Conference Article
Journal/Proceeding Title: European Conference on Computer Vision (ECCV)
Version: Author's manuscript

Items in OAR@Princeton are protected by copyright, with all rights reserved, unless otherwise indicated.