Date:
Fri, 10/17/2014 - 11:00am - 12:30pm
Location:
CCRMA Seminar Rooom
Event Type:
Hearing Seminar
We propose the product-of-filters (PoF) model, a generative model that decomposes audio spectra as sparse linear combinations of "filters" in the log-spectral domain. PoF makes similar assumptions to those used in the classic homomorphic filtering approach to signal processing, but replaces hand-designed decompositions built of basic signal processing operations with a learned decomposition based on statistical inference. When applied to speech, PoF discovers a source-filter representation of speech, despite its lack of any explicit prior knowledge about the mechanisms of vocalization. The PoF model can be used as a prior in more complicated models, permitting applications to problems such as dereverberation and bandwidth expansion.
Bio:
Matt Hoffman is a research scientist in the Creative Technologies Laboratory in Adobe Research. Before that, he was a postdoc working with Prof. Andrew Gelman in the Statistics Department at Columbia University. He did his Ph.D. at Princeton University in Computer Science working in the Sound Lab with Prof. Perry Cook and Prof. David Blei. His research interests include developing efficient Bayesian (and pseudo-Bayesian) inference algorithms; hierarchical probabilistic modeling of audio, text, and marketing data; audio feature extraction, music information retrieval, and the application of music information retrieval and modeling techniques to musical synthesis.