Voice Activity Detection
Definition
Voice activity detection distinguishes spoken segments from silence and other noises in audio recordings.
The temporal detection of segments with speech versus silence or other noises in an audio signal.
Simply explained
The system listens to a recording and marks when someone is speaking and when only pauses or other noises are heard.
Why it matters in practice
The term helps to select, develop, evaluate AI systems in a well-founded manner and realistically assess their technical risks.
Example from business
A transcription solution forwards only detected speech segments and preserves short pauses.
Limits and pitfalls
- Noise, music, whispering, and overlapping speech cause errors; VAD does not identify persons and does not transcribe.
Origin & history
VAD originates from speech coding and telecommunications and was later extended by statistical and neural models.
Use in mid-sized companies
- Possible example: A transcription solution forwards only detected speech segments and preserves short pauses.
Frequently asked questions
What does voice activity detection mean?
The temporal detection of segments with speech versus silence or other noises in an audio signal. What should you watch out for? Noise, music, whispering, and overlapping speech cause errors; VAD does not identify persons and does not transcribe.
Sources
- A statistical model-based voice activity detection. IEEE Signal Processing Letters, 6(1), 1–3. · Jongseo Sohn, Nam Soo Kim und Wonyong Sung (1999): A statistical model-based voice activity detection. IEEE Signal Processing Letters, 6(1), 1–3.
Editorially reviewed on: