Jonathan Vincent

Papers

Total Citations

H-Index

About

Jonathan Vincent is a leading researcher in robot audition and autonomous perception, whose work bridges the gap between artificial hearing and robust visual navigation. His most impactful contribution is the **Open embeddeD Audition System (ODAS)**, a pioneering open-source framework that delivers real-time sound source localization, tracking, and separation for robots and embedded devices. With 23 citations, ODAS has become a foundational tool in the field, enabling machines to perceive their acoustic environment with unprecedented efficiency, while dramatically reducing the computational overhead that previously limited audition on mobile platforms. Vincent also addresses the challenge of dynamic environments in visual SLAM, developing a **fast pipeline for dynamic object tracking and masking** that removes visual features from moving objects to prevent localization drift. This work, cited 7 times, offers a practical, lightweight solution for maintaining mapping accuracy in real-world, cluttered spaces. By integrating auditory and visual perception, Vincent is advancing the next generation of autonomous systems that can hear, see, and navigate intelligently.