Audio transcription converts spoken language from recordings into written text. Modern automatic speech recognition systems also add punctuation, distinguish between speakers and sometimes provide timestamps. It is used for meeting minutes, interviews, subtitles and documenting customer calls. Quality depends heavily on audio quality, accent and specialist vocabulary; for conversations involving personal data, consent and data protection must be considered.
Speech recognition, also called ASR (automatic speech recognition), automatically converts spoken language into written text. Modern systems use neural networks trained on large amounts of audio recordings and cope reasonably well with accents and background noise. Applications include dictation, subtitles, voice assistants and the transcription of meetings and customer calls. Technical terms, proper names and poor audio quality remain common sources of error.
Whisper is an automatic speech recognition model released by OpenAI in September 2022. It converts spoken language in many languages into text and can also translate recordings into English. Its code and model weights are available under the open-source MIT licence, so Whisper can also run on your own hardware, an advantage when audio data should not leave the organisation. Typical uses include transcripts of meetings, interviews and subtitles.
Text-to-speech converts written text into spoken language, making it the counterpart of speech recognition (speech-to-text). Modern systems use neural networks to produce natural-sounding voices with appropriate intonation. TTS is used in telephone announcements, navigation devices, read-aloud functions for accessible websites and voice assistants.
