CaveSeg: Deep Semantic Segmentation and Scene Parsing for Autonomous Underwater Cave Exploration

arXiv:2309.11038v4 Announce Type: replace-cross
Abstract: In this paper, we present CaveSeg - the first visual learning pipeline for semantic segmentation and scene parsing for AUV navigation inside underwater caves. We address the problem of scarce annotated training data by preparing a comprehensive dataset for semantic segmentation of underwater cave scenes. It contains pixel annotations for important navigation markers (e.g. caveline, arrows), obstacles (e.g. ground plain and overhead layers), scuba divers, and open areas for servoing. Through comprehensive benchmark analyses on …

abstract arxiv autonomous cs.cv cs.ro data dataset eess.iv exploration inside navigation paper parsing pipeline segmentation semantic training training data type underwater visual