Joint source-channel modeling and estimation for speech dereverberation

Speech dereverberation is an important challenge in acoustic signal processing because of the detrimental effects upon the signal quality brought by the reverberant components due to acoustic reflections from the walls of the room enclosure. Most of the methods proposed so far such as microphone array beamforming and room impulse response inversion for deconvolution make very little use of the knowledge about the characteristics of the source signal. This paper presents a formulation of speech dereverberation as a probabilistic modeling problem in which joint estimation of the source and the channel (or its inverse) can be accomplished. We discuss reasonable representations of the source prior for inclusion in such a formulation and propose the corresponding solutions.