期刊
IEEE TRANSACTIONS ON IMAGE PROCESSING
卷 30, 期 -, 页码 9442-9455出版社
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/TIP.2021.3126418
关键词
Costs; Reinforcement learning; Deep learning; Three-dimensional displays; Machine learning algorithms; Training; Supervised learning; Real-time stereo matching; disparity estimation; deep learning; reinforcement learning; convolutional neural networks
资金
- National Natural Science Foundation of China [U1933134, U1833128]
- National Key Research and Development Program of China [2018YFC0809500]
- Sichuan Science and Technology Program [2018JY0602, 2018GZDZX0024, 2020YFG0134]
- Sichuan University [2018SCUH0042]
This paper proposes a novel real-time stereo matching method called RLStereo, based on reinforcement learning, which iteratively conducts a few actions to search the disparity value for each pair of stereo images after training. Experimental results demonstrate the high performance of the RLStereo method in terms of speed and accuracy.
Many state-of-the-art stereo matching algorithms based on deep learning have been proposed in recent years, which usually construct a cost volume and adopt cost filtering by a series of 3D convolutions. In essence, the possibility of all the disparities is exhaustively represented in the cost volume, and the estimated disparity holds the maximal possibility. The cost filtering could learn contextual information and reduce mismatches in ill-posed regions. However, this kind of methods has two main disadvantages: 1) cost filtering is very time-consuming, and it is thus difficult to simultaneously satisfy the requirements for both speed and accuracy; 2) thickness of the cost volume determines the disparity range which can be estimated, and the pre-defined disparity range may not meet the demand of practical application. This paper proposes a novel real-time stereo matching method called RLStereo, which is based on reinforcement learning and abandons the cost volume or the routine of exhaustive search. The trained RLStereo makes only a few actions iteratively to search the value of the disparity for each pair of stereo images. Experimental results show the effectiveness of the proposed method, which achieves comparable performances to state-of-the-art algorithms with real-time speed on the public large-scale testset, i.e., Scene Flow.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据