Speech enhancement (current projects)
o Development of MMSE algorithms for speech enhancement.
o Development of noise estimation algorithms suitable for highly non-stationary noisy conditions
o Development of subspace algorithms for non-stationary environments
o
Active noise control for fMRI scanner rooms
[Research supported in part by NIH and done in collaboration with University of Texas Southwestern Medical center (UTSW)]
Speech Compression (past projects)
o Real-time implementation of speech compression algorithms, including VSELP which has been adopted as the North American digital cellular standard, and GSM, the European speech coding standard.
o Design and implementation of vector quantization algorithms for speech coding
Speech Recognition (past projects)
o Development of preprocessing algorithms for digit recognition. Preprocessing algorithms were run real-time on a fixed-point processor.
o Development of alphabet recognition algorithms for name retrieval applications using context-dependent Hidden Markov Models.