The corresponding disparity is estimatedīy matching pixels from rectified image pairs captured by The associate editor coordinating the review of this manuscript andĪpproving it for publication was Naveed Akhtar Stereo depthĮstimation uses the relationship between disparity and depth Stereo depth estimation usually works better. Monocular depth estimation and stereo depth estimation, and Of passive depth estimation, it is widely used in 3D vision.įurthermore, passive depth estimation can be divided into In general, depth estimation can be divided into active depth estimation and passive depth estimation. Therefore, it is crucial for 3D vision to accuratelyĮstimate the missing depth information from images. The depth of information is lost in the process of capturing INDEX TERMS Stereo disparity estimation, 3D convolution, knowledge distillation, compact extractor, costĮstimating indoor and outdoor scenes via images is a challenging problem for 3D vision, which is due to the fact that Have shown that our method achieves competitive performance on the challenging Scene Flow and KITTIīenchmarks while maintaining a very fast running speed. Network and those of the student network, resulting in a more robust distillation process. Furthermore, we present a novelĪdaptive SmoothL1 (ASL) Loss for calculating the similarity between the contextual features of the teacher Transfer contextual features from a teacher network to a student network. The computational cost and maintain the accuracy of disparity, we utilized knowledge distillation scheme to Was proposed to extract the more refined features for constructing multi-scale cost volume. In particular, a compact and efficient multi-scale extractor named MCliqueNet with stacked CliqueBlock In this paper, we proposed an efficient convolutional neural architecture for stereo disparity estimation. Most previous works take advantage of stacked 3D convolutional block to generate fine disparity, but withĪ high computational cost and a large memory consumption. Although manyĮxperimental techniques have been proposed in recent years with the flourishing of deep learning, veryįew studies take into account the optimization of computational complexity and memory consumption. Part by the Science and Technology Program of Fujian Province of China under Grant 2019YZ016006.ĪBSTRACT Stereo disparity estimation is a difficult and crucial task in computer vision. Key Laboratory of Medical Instrumentation and Pharmaceutical Technology, Fuzhou 350108, Chinaģ Imperial Vision Technology, Fuzhou 350002, ChinaĬorresponding author: Tong Tong work was supported in part by the National Natural Science Foundation of China under Grant 61901120 and Grant 61802065, and in Of Physics and Information Engineering, Fuzhou University, Fuzhou 350108, China Received September 4, 2020, accepted October 5, 2020, date of publication October 9, 2020, date of current version November 2, 2020.ĭigital Object Identifier 10.1109/ACCESS.2020.3029832Ĭompact StereoNet: Stereo Disparity Estimation
0 Comments
Leave a Reply. |
Details
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |