Loopy belief propagation Our algorithm, like those of many groups, uses the top - down approach, which tries to model the full sound from
multiple talkers without first picking out special regions in the spectrogram.
To separate the speech of
multiple talkers or to recognize one person's speech, computers represent the sound signal by its spectrum — the energy in the sound at each frequency.
Systematically trying out every possible combination would be very slow, similar to doing the full Viterbi algorithm for
multiple talkers.
«In everyday situations we are frequently confronted with
multiple talkers emitting auditory and visual speech cues, and the brain must decide whether or not to integrate a particular combination of voice and face.»