Source separation is the study of separating signal(s) from mixed sources.
Source separation is the study of separating signal(s) from mixed sources. The goal is to estimate the desired signal and remove other unwanted signals. Source separation methods in digital signal processing are commonly classified as blind and unblind source separation (including semi-blind). In blind source separation, no prior information used to recover target signal. In unblind and semiblind, some prior information is used to estimate target signal such as number of source, mixing matrix and others. The nature of source separation is the ability of human auditory processing to focus on on target signal while ignore other noises in cocktail party phenomena. There are some application of source separation like speech recognition, auralization, noise suppression and signal enhancement and voice assistant.
