The article discusses a study on using incidental sounds produced by people to detect their presence and location around a robot. The researchers collected a dataset of high-quality 4-channel audio paired with 360 RGB data of people moving in different indoor settings. They trained models to predict if there is a moving person nearby and their location using only audio. The method was implemented on a robot, enabling it to track a single person moving quietly using only passive audio sensing. The study highlights the potential of using acoustics for person detection in robotic applications.
Publication date: 6 Oct 2023
Project Page: Not provided
Paper: https://arxiv.org/pdf/2310.03743
