Self-supervised Monocular Depth Estimation for All Day Images using Domain Separation

TL;DR

This paper introduces a domain-separated network for self-supervised monocular depth estimation that effectively handles all-day images by separating private and invariant features, improving performance across varying illumination conditions.

Contribution

The novel approach partitions image information into private and invariant domains, using GAN-generated images and orthogonality constraints to reduce domain gap and enhance depth estimation accuracy.

Findings

Achieves state-of-the-art results on Oxford RobotCar dataset.

Effectively handles day and night image variations.

Improves depth map quality across all-day conditions.

Abstract

Remarkable results have been achieved by DCNN based self-supervised depth estimation approaches. However, most of these approaches can only handle either day-time or night-time images, while their performance degrades for all-day images due to large domain shift and the variation of illumination between day and night images. To relieve these limitations, we propose a domain-separated network for self-supervised depth estimation of all-day images. Specifically, to relieve the negative influence of disturbing terms (illumination, etc.), we partition the information of day and night image pairs into two complementary sub-spaces: private and invariant domains, where the former contains the unique information (illumination, etc.) of day and night images and the latter contains essential shared information (texture, etc.). Meanwhile, to guarantee that the day and night images contain the same information, the domain-separated network takes the day-time images and corresponding night-time images (generated by GAN) as input, and the private and invariant feature extractors are learned by orthogonality and similarity loss, where the domain gap can be alleviated, thus better depth maps can be expected. Meanwhile, the reconstruction and photometric losses are utilized to estimate complementary information and depth maps effectively. Experimental results demonstrate that our approach achieves state-of-the-art depth estimation results for all-day images on the challenging Oxford RobotCar dataset, proving the superiority of our proposed approach.