Robust Enhancement of Reverberant Speech
This paper deals with enhancement of speech recording in a highly reverberant environment, using a single microphone. Motivation for this research is application to forensic investigations, where recording conditions are often suboptimal, with acoustical environments which may be highly reverberant. The paper considers the conventional deconvolutional techniques usually proposed for speech dereverberation and shows why they are unsuccessful in dealing with severe reverberation in practical applications.: Preprint
This paper costs $20 for non-members, $5 for AES members and is free for E-Library subscribers.