文献类型： 外文期刊

作者： Yang, Xiaowei ¹ ; He, Lin ¹ ; Zhao, Yong ⁴ ; Sang, Haiwei ⁵ ; Yang, Zuliu ⁴ ; Cheng, Xianjing ⁶ ;

作者机构： 1.Guizhou Univ, Sch Mech Engn, Guiyang 550025, Peoples R China

2.Guizhou Acad Agr Sci, Guizhou Tea Res Inst, Guiyang 550006, Peoples R China

3.Liupanshui Normal Coll, Liupanshui 553004, Peoples R China

4.Peking Univ, Shenzhen Grad Sch, Sch Elect & Comp Engn, Shenzhen 518055, Peoples R China

5.Guizhou Educ Univ, Sch Math & Big Data, Guiyang 550018, Peoples R China

6.Zunyi Normal Univ, Res Inst Qianbei Informat Technol, Zunyi 563006, Guizhou, Peoples R China

关键词： Feature extraction; Three-dimensional displays; Convolution; Data mining; Image edge detection; Convolutional neural networks; Neural network; stereo matching; multi-scale attention module; feature refinement module; 3D attention aggregation module

期刊名称：IEEE ACCESS （ 影响因子：3.367； 五年影响因子：3.671 ）

ISSN： 2169-3536

年卷期： 2020 年 8 卷

页码：

收录情况： SCI

摘要： In recent years, convolutional neural network (CNN) algorithms promote the development of stereo matching and make great progress, but some mismatches still occur in textureless, occluded and reflective regions. In feature extraction and cost aggregation, CNNs will greatly improve the accuracy of stereo matching by utilizing global context information and high-quality feature representations. In this paper, we design a novel end-to-end stereo matching algorithm named Multi-Attention Network (MAN). To obtain the global context information in detail at the pixel-level, we propose a Multi-Scale Attention Module (MSAM), combining a spatial pyramid module with an attention mechanism, when we extract the image features. In addition, we introduce a feature refinement module (FRM) and a 3D attention aggregation module (3D AAM) during cost aggregation so that the network can extract informative features with high representational ability and high-quality channel attention vectors. Finally, we obtain the final disparity through bilinear interpolation and disparity regression. We evaluate our method on the Scene Flow, KITTI 2012 and KITTI 2015 stereo datasets. The experimental results show that our method achieves state-of-the-art performance and that every component of our network is effective.