METHOD AND APPARATUS FOR MULTI-SENSORY SPEECH ENHANCEMENT

摘要

PROBLEM TO BE SOLVED: To provide a method and a system that use an auxiliary sensor signal received from a sensor other than an air conduction microphone to estimate a clean speech value.SOLUTION: The estimation uses either the auxiliary sensor signal alone, or the auxiliary sensor signal in conjunction with the air conduction microphone signal. The clean speech value is estimated without using a model trained from noisy training data collected from the air conduction microphone. Correction vectors are added to a vector formed from the auxiliary sensor signal in order to form a filter, which is applied to the air conductive microphone signal to produce the clean speech estimate. The pitch of a speech signal is determined from the auxiliary sensor signal and is used to decompose the air conduction microphone signal. The decomposed signal is then used to determine a clean signal estimate.