Submitted by Zachary Teed.

Submission data

Full nameRAFT-Stereo: Multilevel Recurrent Field Transforms for Stereo Matching
DescriptionWe introduce RAFT-Stereo, a new deep architecture for rectified stereo based on the optical flow network RAFT. We introduce multi-level convolutional GRUs, which more efficiently propagate information across the image. A modified version of RAFT-Stereo can perform accurate real-time inference.
Publication titleRAFT-Stereo: Multilevel Recurrent Field Transforms for Stereo Matching
Publication authorsLahav Lipson, Zachary Teed, and Jia Deng
Publication venue3DV
Publication URLhttps://arxiv.org/abs/2109.07547
Programming language(s)Pytthon
HardwareRTX 6000
Source code or download URLhttps://github.com/princeton-vl/RAFT-Stereo
Submission creation date5 Nov, 2020
Last edited4 Oct, 2021

High-res multi-view results



Infoallhigh-res
multi-view
indooroutdoorbotani.boulde.bridgedoorexhibi.lectur.living.loungeobserv.old co.statueterrac.
No results yet.

Low-res many-view results



Infoalllow-res
many-view
indooroutdoorlakesidesand boxstorage roomstorage room 2tunnel
No results yet.

Low-res two-view results



Infoalllakes. 1llakes. 1ssand box 1lsand box 1sstora. room 1lstora. room 1sstora. room 2lstora. room 2sstora. room 2 1lstora. room 2 1sstora. room 2 2lstora. room 2 2sstora. room 3lstora. room 3stunnel 1ltunnel 1stunnel 2ltunnel 2stunnel 3ltunnel 3s
permissivetwo views0.570.260.630.540.360.690.541.220.721.611.500.750.530.640.440.180.150.170.150.170.17

SLAM results



allboxesboxes darkbuddhacables 4cables 5desk 1desk 2desk changing 2desk dark 1desk dark 2desk global light changesdesk ir lightdinodroneforeground occlusionhelmetkidnap 2lamplarge loop 2large loop 3large non loopmotion 2motion 3motion 4planar 1reflective 2scale changetable 1table 2table 5table 6table global light changestable local light changestable scenetrashbin
MethodInfo
No results yet.