Files in this item



application/pdfECE499-Fa2009-bharadwaj.pdf (127kB)Restricted to U of Illinois
(no description provided)PDF


Title:Advances in Sparse Classification
Author(s):Bharadwaj, Sujeeth
Contributor(s):Hasegawa-Johnson, Mark
Subject(s):compressed sensing
speech recognition
nonparametric speech recognition
sparse classification
sparse projection
Abstract:A recent result in compressed sensing (CS) allows us to perform non-parametric speech recognition that is robust to noise, and that requires few training examples. By taking fixed length representations of training samples and stacking them in a matrix, we form a frame, or an over-complete basis. Gemmeke and Cranen have shown that sparse projections onto this frame recover the correct transcription with 91% accuracy at -5 dB SNR. We propose that the goal of speech recognition is not sparse projection onto training tokens, but onto training types. Sparse projection onto types can be achieved by building a frame for each word in the dictionary, and stacking the frames to form a rank 3 tensor. Speech recognition is performed by convex linear projection onto the tensor, with sparsity enforced only in the index that specifies type. We derive a mixed L1/L2 relaxation that can be globally optimized using Newton descent.
Issue Date:2009-12
Publication Status:unpublished
Peer Reviewed:not peer reviewed
Date Available in IDEALS:2014-01-22

This item appears in the following Collection(s)

Item Statistics