Learning Depth from Monocular Videos Using Synthetic Data: A Temporally-Consistent Domain Adaptation Approach

TL;DR

This paper introduces a novel domain adaptation approach that leverages synthetic videos with ground-truth depth labels to improve monocular depth estimation from real videos, addressing the style gap and temporal consistency issues.

Contribution

It proposes a temporally-consistent domain adaptation method that uses synthetic data with ground-truth labels and temporal constraints to enhance depth prediction in real videos.

Findings

Achieves comparable performance to state-of-the-art methods.

Effectively filters moving regions using learned moving masks.

Utilizes synthetic data to reduce reliance on expensive real-world labels.

Abstract

Majority of state-of-the-art monocular depth estimation methods are supervised learning approaches. The success of such approaches heavily depends on the high-quality depth labels which are expensive to obtain. Some recent methods try to learn depth networks by leveraging unsupervised cues from monocular videos which are easier to acquire but less reliable. In this paper, we propose to resolve this dilemma by transferring knowledge from synthetic videos with easily obtainable ground-truth depth labels. Due to the stylish difference between synthetic and real images, we propose a temporally-consistent domain adaptation (TCDA) approach that simultaneously explores labels in the synthetic domain and temporal constraints in the videos to improve style transfer and depth prediction. Furthermore, we make use of the ground-truth optical flow and pose information in the synthetic data to learn moving mask and pose prediction networks. The learned moving masks can filter out moving regions that produces erroneous temporal constraints and the estimated poses provide better initializations for estimating temporal constraints. Experimental results demonstrate the effectiveness of our method and comparable performance against state-of-the-art.