首页|On Robust Cross-view Consistency in Self-supervised Monocular Depth Estimation
On Robust Cross-view Consistency in Self-supervised Monocular Depth Estimation
扫码查看
点击上方二维码区域，可以放大扫码查看
原文链接
NETL
NSTL
万方数据
Remarkable progress has been made in self-supervised monocular depth estimation(SS-MDE)by exploring cross-view con-sistency,e.g.,photometric consistency and 3D point cloud consistency.However,they are very vulnerable to illumination variance,oc-clusions,texture-less regions,as well as moving objects,making them not robust enough to deal with various scenes.To address this challenge,we study two kinds of robust cross-view consistency in this paper.Firstly,the spatial offset field between adjacent frames is obtained by reconstructing the reference frame from its neighbors via deformable alignment,which is used to align the temporal depth features via a depth feature alignment(DFA)loss.Secondly,the 3D point clouds of each reference frame and its nearby frames are calcu-lated and transformed into voxel space,where the point density in each voxel is calculated and aligned via a voxel density alignment(VDA)loss.In this way,we exploit the temporal coherence in both depth feature space and 3D voxel space for SS-MDE,shifting the"point-to-point"alignment paradigm to the"region-to-region"one.Compared with the photometric consistency loss as well as the rigid point cloud alignment loss,the proposed DFA and VDA losses are more robust owing to the strong representation power of deep fea-tures as well as the high tolerance of voxel density to the aforementioned challenges.Experimental results on several outdoor bench-marks show that our method outperforms current state-of-the-art techniques.Extensive ablation study and analysis validate the effect-iveness of the proposed losses,especially in challenging scenes.The code and models are available at https://github.com/sunnyHelen/RCVC-depth.
3D visiondepth estimationcross-view consistencyself-supervised learningmonocular perception
Haimei Zhao、Jing Zhang、Zhuo Chen、Bo Yuan、Dacheng Tao
展开 >
School of Computer Science,University of Sydney,Sydney 2008,Australia
Shenzhen International Graduate School,Tsinghua University,Shenzhen 518055,China
School of Information Technology & Electrical Engineering,University of Queensland,Brisbane 4072,Australia
NETL
NSTL
万方数据
