Abstract:Domain shift, where deviations between training and deployment data distributions degrade model performance, is a key challenge in underwater environments. Existing benchmarks testing performance for underwater domain shift simulate variability through synthetic style transfer. This fails to capture intrinsic scene factors such as visibility, illumination, scene composition, or acquisition factors, limiting analysis of real-world effects. We propose a labeling framework that defines underwater domains using measurable image, scene, and acquisition characteristics. Unlike prior benchmarks, it captures physically meaningful factors, enabling semantically consistent image grouping and supporting domain-specific evaluation of detection performance including failure analysis. We validate this on public datasets, showing systematic variations across domain factors and revealing hidden failure modes.
| Comments: | Poster Presentation at ICRA 2026 Workshop S2S |
| Subjects: | Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG); Robotics (cs.RO) |
| Cite as: | arXiv:2604.26174 [cs.CV] |
| (or arXiv:2604.26174v1 [cs.CV] for this version) | |
| https://doi.org/10.48550/arXiv.2604.26174 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Melanie Wille [view email]
[v1]
Tue, 28 Apr 2026 23:28:14 UTC (1,794 KB)