Depth Completion with Twin Surface Extrapolation at Occlusion Boundaries

TL;DR

This paper introduces a novel depth completion method that models occlusion boundaries with twin-surface extrapolation, improving depth accuracy by explicitly handling foreground and background regions, validated across multiple datasets.

Contribution

It proposes a multi-hypothesis twin-surface extrapolation approach with an asymmetric loss function for better depth completion at occlusion boundaries.

Findings

Outperforms state-of-the-art methods on KITTI, NYU2, and Virtual KITTI datasets.

Effectively models occlusion boundaries with twin-surface extrapolation.

Demonstrates improved depth accuracy and boundary sharpness.

Abstract

Depth completion starts from a sparse set of known depth values and estimates the unknown depths for the remaining image pixels. Most methods model this as depth interpolation and erroneously interpolate depth pixels into the empty space between spatially distinct objects, resulting in depth-smearing across occlusion boundaries. Here we propose a multi-hypothesis depth representation that explicitly models both foreground and background depths in the difficult occlusion-boundary regions. Our method can be thought of as performing twin-surface extrapolation, rather than interpolation, in these regions. Next our method fuses these extrapolated surfaces into a single depth image leveraging the image data. Key to our method is the use of an asymmetric loss function that operates on a novel twin-surface representation. This enables us to train a network to simultaneously do surface extrapolation and surface fusion. We characterize our loss function and compare with other common losses. Finally, we validate our method on three different datasets; KITTI, an outdoor real-world dataset, NYU2, indoor real-world depth dataset and Virtual KITTI, a photo-realistic synthetic dataset with dense groundtruth, and demonstrate improvement over the state of the art.