Reveal of Domain Effect: How Visual Restoration Contributes to Object Detection in Aquatic Scenes

TL;DR

This paper investigates how visual restoration affects object detection in underwater scenes, revealing that restoration reduces domain shift and improves detection in real-world conditions, with specific insights into domain quality and generalization.

Contribution

It uncovers the specific role of visual restoration in mitigating domain effects and enhancing underwater object detection in real-world scenarios.

Findings

Domain quality has minimal impact on within-domain detection accuracy.

Low-quality domains improve cross-domain generalization.

Restoration reduces domain shift, aiding detection in the wild.

Abstract

Underwater robotic perception usually requires visual restoration and object detection, both of which have been studied for many years. Meanwhile, data domain has a huge impact on modern data-driven leaning process. However, exactly indicating domain effect, the relation between restoration and detection remains unclear. In this paper, we generally investigate the relation of quality-diverse data domain to detection performance. In the meantime, we unveil how visual restoration contributes to object detection in real-world underwater scenes. According to our analysis, five key discoveries are reported: 1) Domain quality has an ignorable effect on within-domain convolutional representation and detection accuracy; 2) low-quality domain leads to higher generalization ability in cross-domain detection; 3) low-quality domain can hardly be well learned in a domain-mixed learning process; 4) degrading recall efficiency, restoration cannot improve within-domain detection accuracy; 5) visual restoration is beneficial to detection in the wild by reducing the domain shift between training data and real-world scenes. Finally, as an illustrative example, we successfully perform underwater object detection with an aquatic robot.