Adaptive Reverberation Absorption using Non-stationary Masking Components Detection for Intelligibility Improvement
This work addresses speech intelligibility enhancement for applications in noisy-reverberant environments, representing an incremental improvement over existing methods.
The paper tackled the problem of improving speech intelligibility in noisy-reverberant conditions by proposing a time domain absorption approach that detects masking distortions using non-stationarity, without requiring prior knowledge of speech statistics or room information. The results showed higher intelligibility improvement compared to competing methods, as measured by three intelligibility metrics and a perceptual listening test.
This letter proposes a new time domain absorption approach designed to reduce masking components of speech signals under noisy-reverberant conditions. In this method, the non-stationarity of corrupted signal segments is used to detect masking distortions based on a defined threshold. The nonstationarity is objectively measured and is also adopted to determine the absorption procedure. Additionally, no prior knowledge of speech statistics or of the room information is required for this technique. Three intelligibility measures (ESII, ASIIST, SRMRnorm) and a perceptual listening test are used for evaluation. The experiments results show that the proposed scheme leads to a higher intelligibility improvement when compared to competing methods.