Missing data imputation using compressive sensing techniques for connected digit recognition

An effective way to increase the noise robustness of automatic speech recognition is to label noisy speech features as either reliable or unreliable (missing) prior to decoding, and to replace the missing ones by clean speech estimates. We present a novel method based on techniques from the field of Compressive Sensing to obtain these clean speech estimates. Unlike previous imputation frameworks which work on a frame-by-frame basis, our method focuses on exploiting information from a large time-context. Using a sliding window approach, denoised speech representations are constructed using a sparse representation of the reliable features in an overcomplete dictionary of clean, fixed-length speech exemplars. We demonstrate the potential of our approach with experiments on the AURORA-2 connected digit database.