Title: Sound Source Localization and Its Applications for Robots
Type and Duration: Half-day workshop at ICRA 19
Acoustic signals are ubiquitous, and both humans and animals utilize them for understanding surrounding environments and for performing various interactions. However, there is relatively little work in robotics on the use of acoustic signals for scene understanding or analysis. There has been some work over the last two decades on robot audition and machine hearing, the goal of which is to improve human-robot interactions and scene navigation. Thanks to recent advances in deep learning and acoustic sensing, robots and computers have been improved considerably in terms of human speech understanding and interpreting other acoustic signals. However, these developments are mainly limited to direct sound signals in simple settings (e.g., a living room) or near-field speech recognition.
In this context, we will cover the following topics:
- Interactive sound propagation technologies Localization applications for drone and underwater robots
- Deep learning techniques for multi-party interactions and simultaneous talkers
- Real-time localization techniques considering non line of sight wave effects (diffraction)
- Joint optimizations between SLAM and acoustic approaches.
- Audio-motor localization
These topics are covered by top-notch researchers in the related fields and these speakers are actively working in the field of robotics; they are regularly participating and publishing at ICRA. We have already invited related speakers and interacted with them to set the scope of this workshop.
Our invited speakers who confirmed the participation for the workshop include:
- Dinesh Manocha, IEEE/ACM/AAAI Fellow, Professor at Univ. Maryland, USA
- Prof. Hiroshi G. Okuno, IEEE Fellow, Waseda Univ., Japan
- Frank Dellaert, Professor, Georgia Tech (formerly at Facebook), USA
- Sung-eui Yoon, IEEE/ACM Senior member, KAIST, Korea
- Mandar Chitre, National University of Singapore, Editor-in-Chief for the IEEE Journal of Oceanic Engineering, Singapore
- Patrick Danes, Professor at Université de Toulouse and LAAS-CNRS, France
- Jean-Marc Odobez, Professor at Maine University and Idiap Research Institute/EPFL, France
- Kazuhiro Nakadai, IEEE Senior member, Principal research at Honda Research Institute, and Professor at Tokyo Institute of Technology, Japan