Voice activity detection

Voice activity detection (VAD), also known as speech activity detection or speech detection, is the detection of the presence or absence of human speech, used in <a href="/facts/Speech_processing/jUjXoYwc">speech processing</a>. The main uses of VAD are in <a href="/facts/Speaker_diarization/DdcRETMS">speaker diarization</a>, <a href="/facts/Speech_coding/xs5gvMXq">speech coding</a> and <a href="/facts/Speech_recognition/z7S7Pgk6">speech recognition</a>. It can facilitate speech processing, and can also be used to deactivate some processes during non-speech section of an audio session: it can avoid <a href="/facts/Silence_compression/F7XvUdSc">unnecessary coding</a>/transmission of silence <a href="/facts/Packet_switching/olDlDP1R">packets</a> in <a href="/facts/Voice_over_Internet_Protocol/5QGBbXYg">Voice over Internet Protocol</a> (VoIP) applications, saving on computation and on <a href="/facts/Bandwidth_(computing)/lY6Tb4gX">network bandwidth</a>.
VAD is an important enabling technology for a variety of speech-based applications. Therefore, various VAD algorithms have been developed that provide varying features and compromises between <a href="/facts/Latency_(engineering)/6Bjhn30H">latency</a>, <a href="/facts/Sensitivity_(electronics)/1SAUBAss">sensitivity</a>, accuracy and computational cost. Some VAD algorithms also provide further analysis, for example whether the speech is <a href="/facts/Voice_(phonetics)/VdqrEGQ3">voiced</a>, unvoiced or <a href="/facts/Sustain/FJylhHed">sustained</a>. Voice activity detection is usually independent of language.
It was first investigated for use on <a href="/facts/Time-assignment_speech_interpolation/KReWH9Nl">time-assignment speech interpolation</a> (TASI) systems.

Voice activity detection open-in-new

Voice activity detection