Single-channel noise reduction exploits speech’s temporal, spectral, and statistical properties

Evidence

In addition to directional microphone systems, the SNR in the hearing aid output can be increased by single-channel noise reduction schemes (Dillon, 2012). These techniques use the temporal, spectral, and statistical information of the incoming sound to suppress the noise. Various speech enhancement algorithms assume some properties for speech and use these properties to attenuate unwanted noise and preserve speech-like signals. — Ville Pulkki, Matti Karjalainen, Communication Acoustics- An…, 2752

Categorization

• XREF: Connects to statistical speech enhancement and spectral subtraction methods used in telephony and voice assistants.