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:
Full metadata record
DC FieldValueLanguage
dc.contributor.authorTeed, Zachary-
dc.contributor.authorDeng, Jia-
dc.identifier.citationTeed, 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_24en_US
dc.description.abstractWe 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
dc.format.extent402 - 419en_US
dc.relation.ispartofEuropean Conference on Computer Vision (ECCV)en_US
dc.rightsAuthor's manuscripten_US
dc.titleRAFT: Recurrent All-Pairs Field Transforms for Optical Flowen_US
dc.typeConference Articleen_US

Files in This Item:
File Description SizeFormat 
RecurrentAllPairsFieldTransformsOpticalFlow.pdf7.2 MBAdobe PDFView/Download

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