| Biography | |
|---|---|
 Prof. Wenwu Wang University of Surrey, UK |
|
| Title: Dictionary Learning and Sparse Signal Recovery for Nonlinear Compressive Measurements | |
| Abstract: Sparse representations and dictionary learning have been used widely in linear inverse problems, such as denoising, inpainting, deblurring, or super-resolution. However, they have been less explored for nonlinear measurements. In this talk, we present a new method for signal recovery and dictionary learning from nonlinear measurements, such as clipping (also called saturation), and quantization. Different from conventional methods, where recovering a signal from clipped and quantized measurements is often formulated as a constrained optimization problem, we propose a unified framework for signal recovery from clipped, quantized, as well as linear measurements. With a data-fidelity term that promotes consistency with the nonlinear measurement function, we generalize the linear least-squares loss function commonly used in sparse decompositions, and show that under some conditions on the measurement function, the proposed loss is convex, and continuously differentiable with a closed-form gradient, which makes it suitable for a range of optimization algorithms. This allows us to extend classical sparse decomposition algorithms to deal with nonlinear measurements. We then discuss how to learn a dictionary from the nonlinear compressive measurements, and demonstrate its improved performance for signal reconstruction, over the use of fixed dictionaries. | |
| Biography: Wenwu Wang is a Professor in Signal Processing and Machine Learning, and a Co-Director of the Machine Audition Lab within the Centre for Vision Speech and Signal Processing, University of Surrey, UK. His current research interests include audio-visual signal processing, machine learning and perception, and machine audition (listening). He has (co)-authored over 250 publications in these areas. He is a (co-)recipient of over 15 awards including the Judge’s Award on DCASE 2020, the Reproducible System Award on DCASE 2019 and 2020, Best Student Paper Award on LVA/ICA 2018, the Best Oral Presentation on FSDM 2016, the TVB Europe Award for Best Achievement in Sound in 2016, and the Best Solution Award on the Dstl Challenge in 2012. He is a Senior Area Editor for IEEE Transactions on Signal Processing, an Associate Editor for IEEE/ACM Transactions on Audio Speech and Language Processing, an Associate Editor for EURASIP Journal on Audio Speech and Music Processing. He is a Specialty Editor in Chief for Frontiers in Signal Processing. He is an Elected Member of the IEEE SPTM Technical Committee, and IEEE MLSP Technical Committee, and International Steering Committee of Latent Variable Analysis and Signal Separation. He was a Publication Co-Chair for ICASSP 2019, and a Satellite Workshop Co-Chair for INTERSPEECH 2022. | |
