Pseudo-Label Guided Object Detection in Sparsely Annotated Underwater Optical Images

Object detection in underwater optical imagery plays a crucial role in various fields related to underwater exploration. However, manual annotation of such images often results in incomplete ground truth due to its severe degradation. In this study, we address the issue of incomplete supervision signals in degraded underwater images by reframing it as a sparse annotation challenge. Specifically, we present a novel method for object detection in sparsely annotated underwater scenarios. Our approach involves an effective pseudo-label gen eration network designed to produce labels for the degraded foreground lacking annotations. To mitigate potential back ground noise resulting from the discrepancy between the fixed confidence threshold and its dynamic distribution, we intro duce a novel dynamic adaptive confidence threshold (DACT) method. In addition, a novel adaptive geometric prior-based noise reduction (AGPNR) strategy is designed to eliminate noisy pseudo-labels with low-quality localization. We validate and analyze our approach through experiments on publicly available underwater optical image datasets. The results demonstrate that our approach achieves significant performance improve ments across various sparsity conditions. Compared with existing state-of-the-art models, our proposed approach delivers sig nificantly superior average precision (AP) performance while maintaining fast inference speeds.