Deep Domain Adversarial Adaptation for Photon-efficient Imaging

TL;DR

This paper introduces a domain adversarial adaptation method for photon-efficient LiDAR imaging, improving real-world performance by leveraging unlabeled data and addressing domain shift issues.

Contribution

It presents a novel domain adaptation approach that enhances photon-efficient imaging models' robustness to real-world conditions without requiring labeled data.

Findings

Outperforms existing methods on simulated and real data

Reduces performance degradation in realistic scenarios

Efficiently utilizes unlabeled real-world data

Abstract

Photon-efficient imaging with the single-photon light detection and ranging (LiDAR) captures the three-dimensional (3D) structure of a scene by only a few detected signal photons per pixel. However, the existing computational methods for photon-efficient imaging are pre-tuned on a restricted scenario or trained on simulated datasets. When applied to realistic scenarios whose signal-to-background ratios (SBR) and other hardware-specific properties differ from those of the original task, the model performance often significantly deteriorates. In this paper, we present a domain adversarial adaptation design to alleviate this domain shift problem by exploiting unlabeled real-world data, with significant resource savings. This method demonstrates superior performance on simulated and real-world experiments using our home-built up-conversion single-photon imaging system, which provides an efficient approach to bypass the lack of ground-truth depth information in implementing computational imaging algorithms for realistic applications.