Speech Dereverberation in the STFT Domain
This addresses the problem of speech quality and intelligibility degradation due to reverberation for applications like audio processing, though it appears incremental as it builds on known linear filtering approaches.
The authors tackled speech dereverberation by proposing a novel single-channel method using a linear filter in the STFT domain, which can resolve any reverberant signal to a non-reverberant signal given knowledge of the impulse response.
Reverberation is damaging to both the quality and the intelligibility of a speech signal. We propose a novel single-channel method of dereverberation based on a linear filter in the Short Time Fourier Transform domain. Each enhanced frame is constructed from a linear sum of nearby frames based on the channel impulse response. The results show that the method can resolve any reverberant signal with knowledge of the impulse response to a non-reverberant signal.