Martin Aubard
Papers
5
Total Citations
62
H-Index
3
About
Martin Aubard is a leading researcher in underwater robotics, specializing in the integration of deep learning for autonomous underwater vehicles (AUVs). His work focuses on overcoming the unique challenges of sonar-based perception, real-time environmental interaction, and mission autonomy in subsea environments. Aubard’s most influential contribution is his comprehensive 2025 overview on sonar-based deep learning, which has garnered 33 citations and established a foundational framework for robust underwater AI. He developed a real-time automatic wall detection and localization system using side-scan sonar images (17 citations), significantly advancing AUV navigation in uncertain environments. His innovative application of knowledge distillation to YOLOX-ViT for object detection (6 citations) demonstrates how to reduce model size without sacrificing performance, enabling efficient onboard processing. Aubard also contributed to the LSTS Toolchain for deploying deep learning on AUVs and created a behavior-tree-based framework for pipeline inspection mission planning and safety assessment. His work bridges the gap between theoretical AI advances and practical underwater deployment, making him a key figure in the push toward fully autonomous underwater operations.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3Knowledge Distillation in YOLOX-ViT for Side-Scan Sonar Object Detection6 citations · 2024
- 4
- 5