[CVPR 2022 Oral] MaGNet

[CVPR 2022 Oral] Multi-View Depth Estimation by Fusing Single-View Depth Probability with Multi-View Geometry

by Gwangbin Bae, Ignas Budvytis, and Roberto Cipolla
University of Cambridge

arXiv: https://arxiv.org/abs/2112.08177
code: https://github.com/baegwangbin/MaGNet

Please feel free to contact me (gb585@cam.ac.uk) if you have any questions!

-

Abstract

We present MaGNet (Monocular and Geometric Network), a novel framework for fusing single-view depth probability with multi-view geometry, to improve the accuracy, robustness and efficiency of multi-view depth estimation. For each frame, MaGNet estimates a single-view depth probability distribution, parameterized as a pixel-wise Gaussian. The distribution estimated for the reference frame is then used to sample per-pixel depth candidates. Such probabilistic sampling enables the network to achieve higher accuracy while evaluating fewer depth candidates. We also propose depth consistency weighting for the multi-view matching score, to ensure that the multi-view depth is consistent with the single-view predictions. The proposed method achieves state-of-the-art performance on ScanNet, 7-Scenes and KITTI. Qualitative evaluation demonstrates that our method is more robust against challenging artifacts such as texture-less/reflective surfaces and moving objects.

Видео [CVPR 2022 Oral] MaGNet канала Gwangbin Bae