典型文献
Multilevel Disparity Reconstruction Network for Real-Time Stereo Matching
文献摘要：
Recently,stereo matching algorithms based on end-to-end convolutional neural networks achieve excel-lent performance far exceeding traditional algorithms.Current state-of-the-art stereo matching networks mostly rely on full cost volume and 3D convolutions to regress dense disparity maps.These modules are computationally complex and high consumption of memory,and difficult to deploy in real-time applications.To overcome this problem,we propose multilevel disparity reconstruction network,MDRNet,a lightweight stereo matching network without any 3D convolutions.We use stacked residual pyramids to gradually reconstruct disparity maps from low-level resolution to full-level resolution,replacing common 3D computation and optimization convolutions.Our approach achieves a competitive performance compared with other algorithms on stereo benchmarks and real-time inference at 30 frames per second with 4×104 resolutions.
文献关键词：
中图分类号：
作者姓名：
LIU Zhuoran;ZHAO Xu
作者机构：
School of Electronic Information and Electrical Engineering,Shanghai Jiao Tong University,Shanghai 200240,China
文献出处：
引用格式：
[1]LIU Zhuoran;ZHAO Xu-.Multilevel Disparity Reconstruction Network for Real-Time Stereo Matching)[J].上海交通大学学报（英文版）,2022(05):715-722
A类：
Disparity,MDRNet
B类：
Multilevel,Reconstruction,Network,Real,Time,Stereo,Matching,Recently,stereo,matching,algorithms,end,convolutional,neural,networks,excel,lent,performance,far,exceeding,traditional,Current,state,art,mostly,rely,full,cost,volume,convolutions,regress,dense,disparity,maps,These,modules,computationally,complex,high,consumption,memory,difficult,deploy,real,applications,To,overcome,this,problem,propose,multilevel,reconstruction,lightweight,without,any,We,use,stacked,residual,pyramids,gradually,from,low,replacing,common,optimization,Our,approach,achieves,competitive,compared,other,benchmarks,inference,frames,second,resolutions
AB值：
0.628903
相似文献
机标中图分类号，由域田数据科技根据网络公开资料自动分析生成，仅供学习研究参考。
