Underwater Object Detection using Transfer Learning with Deep Learning
Zhu Kaiyan, Xiang Li, Song Weibo · 2020
In recent years, Deep learning based methods have proven their excellent performance in generic object detection. However, underwater object detection is still a challenge, especially from underwater optical images. In contrast to generic datasets, underwater images usually have color shift and low contrast, underwater detection datasets are scarce and the objects in the available underwater datasets and real applications are usually small. To address these issues, we propose a underwater object detection algorithm which choose Yolov3-tiny network as backbone and pre-train on Pascal VOC datasets. Experiments show that our proposed method improves the performance of region-based object detectors on CHINAMM2019 datasets, compared with SSD algorithm and Yolov3.