Our new X account is live! Follow @wizwand_team for updates
WorkDL logo mark

Bilateral Grid Learning for Stereo Matching Networks

About

Real-time performance of stereo matching networks is important for many applications, such as automatic driving, robot navigation and augmented reality (AR). Although significant progress has been made in stereo matching networks in recent years, it is still challenging to balance real-time performance and accuracy. In this paper, we present a novel edge-preserving cost volume upsampling module based on the slicing operation in the learned bilateral grid. The slicing layer is parameter-free, which allows us to obtain a high quality cost volume of high resolution from a low-resolution cost volume under the guide of the learned guidance map efficiently. The proposed cost volume upsampling module can be seamlessly embedded into many existing stereo matching networks, such as GCNet, PSMNet, and GANet. The resulting networks are accelerated several times while maintaining comparable accuracy. Furthermore, we design a real-time network (named BGNet) based on this module, which outperforms existing published real-time deep stereo matching networks, as well as some complex networks on the KITTI stereo datasets. The code is available at https://github.com/YuhuaXu/BGNet.

Bin Xu, Yuhua Xu, Xiaoli Yang, Wei Jia, Yulan Guo• 2021

Related benchmarks

TaskDatasetResultRank
Stereo MatchingKITTI 2015 (test)
D1 Error (Overall)2.19
144
Stereo MatchingKITTI 2015
D1 Error (All)2.19
118
Stereo MatchingKITTI 2012
Error Rate (3px, Noc)1.62
81
Disparity EstimationKITTI 2015 (test)
D1 Error (bg, all)1.91
77
Stereo MatchingKITTI 2012 (test)
Outlier Rate (3px, Noc)1.62
76
Stereo MatchingScene Flow (test)
EPE1.17
70
Stereo MatchingScene Flow
EPE (px)1.17
40
Stereo MatchingMiddlebury 2014 half resolution (generalization)
Bad 2.0 Error11.6
10
Showing 8 of 8 rows

Other info

Code

Follow for update