DCVSMNet: Double Cost Volume Stereo Matching Network
Mahmoud Tahmasebi, Saif Huq, Kevin Meehan, Marion McAfee
Abstract
We introduce the Double Cost Volume Stereo Matching Network (DCVSMNet 1 1 The source code is available at https://github.com/M2219/DCVSMNet . ), a novel architecture characterized by two upper (group-wise correlation) and lower (norm correlation) small cost volumes. Each cost volume is processed separately, and a coupling module is proposed to fuse the geometry information extracted from the upper and lower cost volumes. DCVSMNet is a fast stereo matching network with a 67 ms inference time and strong generalization ability which can produce competitive results compared to state-of-the-art methods. The results on several benchmark datasets show that DCVSMNet achieves better accuracy than methods such as CGI-Stereo and BGNet at the cost of greater inference time.