TUTORIAL 4 INTRO: DeeperSense: AI Methods for Underwater Visual-Acoustic Perception
The applicability of autonomous robots in marine applications relies heavily on their sensing capabilities. Only if a robot is able to perceive its environment even under harsh environmental conditions (low light conditions, turbid and back scatter, etc.), it is able to operate reliably and without putting the mission’s success at risk. The unique challenges posed by the underwater environment necessitate the development and application of advanced Artificial Intelligence (AI) methods for accurate and efficient perception. In this tutorial, a showcase of several AI techniques and algorithms developed within the EU project DeeperSense will be demonstrated. This will include methods for acoustic-to-visual image translation, matching of objects between acoustic and camera views, and segmentation of different sea bed types in side-scan sonar images.