The Third IASTED African Conference on
Modelling and Simulation
Science and Technology Applications for Health and Sustainable Development
September 6 – 8, 2010
A Multimodal Approach to Blind Source Separation of Moving Sources with Applications
The late Professor Colin Cherry, of the Department of Electrical and Electronic Engineering at Imperial College in the UK, is famous for describing the ‘cocktail party problem' in 1953. This problem corresponds to mimicking within a machine the human ability of following one conversation in the presence of other conversations and background noise. The work presented in this talk will be a step towards solving the machine cocktail party problem, drawing on the psychological findings that human auditory perception is often multimodal, i.e., visual cues are fundamental to the process. In particular, blind source separation (BSS) of moving sources will be introduced. The challenge of BSS for moving sources is that the mixing filters are time-varying; thus, the unmixing filters should also be time-varying. However, these are difficult to calculate in real time.
In the proposed approach, the visual modality is used to facilitate the separation for both stationary and moving sources. The movement of the sources is detected by a 3-D tracker based on video cameras. The tracker obtains the sources' positions and velocities based on a Markov Chain Monte Carlo particle filter, which results in a high sampling efficiency. The full BSS solution is formed by integrating a frequency-domain BSS algorithm and beamforming. Experimental results confirm that by using the visual modality, the proposed algorithm not only improves the performance of the BSS algorithm and mitigates the permutation problem for stationary sources, but also provides an acceptable performance for moving sources in a low reverberant environment. The final section of this talk will present applications of this technology in next-generation healthcare, such as in patient monitoring within a smart room environment.
Biography of the Keynote Speaker
Jonathon Chambers is Professor of Communications and Signal Processing within the Department of Electronic and Electrical Engineering at Loughborough University, where he also leads the Advanced Signal Processing Group and is the Deputy Head of Department, Research. He has served as an Associate Editor for IEEE transactions for more than ten years and is currently a member of the IEEE Signal Processing Theory and Methods Technical Committee. He has served as the Technical Programme Chair for the IEEE Workshop on Statistical Signal Processing 2009, Cardiff, and is the Technical Programme Co-Chair for the flagship conference of the IEEE Signal Processing Society ICASSP 2011, Prague. He has supervised about 50 PhD students and has authored or co-authored two research monographs and more than 300 journal and conference papers.