Video Depth Estimation by Fusing Flow-to-Depth Proposals

TL;DR

This paper introduces a novel differentiable flow-to-depth layer for monocular video depth estimation, combining optical flow, camera pose refinement, and depth fusion to produce accurate 3D depth maps with strong cross-dataset generalization.

Contribution

The paper proposes a differentiable flow-to-depth layer integrated with pose refinement and depth fusion, enhancing monocular video depth estimation accuracy and generalization.

Findings

Outperforms state-of-the-art methods on three public datasets.

Demonstrates good cross-dataset generalization from KITTI to Waymo.

Effective integration of flow, pose refinement, and depth fusion improves results.

Abstract

Depth from a monocular video can enable billions of devices and robots with a single camera to see the world in 3D. In this paper, we present an approach with a differentiable flow-to-depth layer for video depth estimation. The model consists of a flow-to-depth layer, a camera pose refinement module, and a depth fusion network. Given optical flow and camera pose, our flow-to-depth layer generates depth proposals and the corresponding confidence maps by explicitly solving an epipolar geometry optimization problem. Our flow-to-depth layer is differentiable, and thus we can refine camera poses by maximizing the aggregated confidence in the camera pose refinement module. Our depth fusion network can utilize depth proposals and their confidence maps inferred from different adjacent frames to produce the final depth map. Furthermore, the depth fusion network can additionally take the depth proposals generated by other methods to improve the results further. The experiments on three public datasets show that our approach outperforms state-of-the-art depth estimation methods, and has reasonable cross dataset generalization capability: our model trained on KITTI still performs well on the unseen Waymo dataset.