InStereo2K:a large real dataset for stereo matching in indoor scenes  

作　　者：Wei BAO Wei WANG Yuhua XU Yulan GUO Siyu HONG Xiaohu ZHANG 

机构地区：[1]School of Electrical Engineering and Automation,Hefei University of Technology,Hefei 230009,China [2]Orbbec Research,Shenzhen 518052,China [3]School of Electronics and Communication Engineering,Sun Yat-sen University,Guangzhou 510006,China [4]College of Electronic Science and Technology,National University of Defense Technology,Changsha 410073,China [5]School of Aeronautics and Astronautics,Sun Yat-sen University,Guangzhou 510006,China

出　　处：《Science China(Information Sciences)》2020年第11期128-138,共11页中国科学（信息科学）（英文版）

基　　金：supported by National Natural Science Foundation of China(Grant Nos.61402489,61972435,61972435,61602499);Natural Science Foundation of Guangdong Province(Grant No.2019A1515011271);Fundamental Research Funds for the Central Universities(Grant No.18lgzd06);Shenzhen Technology and Innovation Committee(Grant No.201908073000399)。

摘　　要：Deep neural networks have shown great success in stereo matching in recent years.On the KITTI datasets,most top performing methods are based on neural networks.However,on the Middlebury datasets,these methods usually do not perform well.The KITTI datasets are collected in outdoor scenes while the Middlebury datasets are collected in indoor scenes.It is commonly believed that the community still lacks a large labelled dataset for stereo matching in indoor scenes.In this paper,we introduce a new stereo dataset called InS tereo2K.It contains 2050 pairs of stereo images with highly accurate groundtruth disparity maps,including 2000 pairs for training and 50 pairs for test.Experimental results show that our dataset can significantly improve the performance of several latest networks(including StereoNet and PSMNet)on the Middlebury 2014 dataset.The large scale,high accuracy and rich diversity of the proposed InS tereo2K dataset provide new opportunities to researchers in the area of stereo matching and beyond.It also takes end-to-end stereo matching methods a step towards practical applications.

关 键 词：stereo matching depth estimation convolutional neural network DATASET 

分 类 号：TP311.13[自动化与计算机技术—计算机软件与理论]