Speech recognition (ASR)
Technology that turns spoken words into text, also known as ASR (automatic speech recognition) or speech-to-text.
Automatic speech recognition (ASR), or speech-to-text, is the technology that listens to audio and writes down what’s said. Modern systems use neural networks trained on thousands of hours of speech and text: they take the sound in small pieces and predict the most likely words, punctuation included.
Don’t confuse it with speaker recognition, which identifies who a voice belongs to. Speech recognition is about what is said.
Its accuracy depends on the language and accent, the vocabulary (unusual proper names are what fails most), noise and people talking at once. That’s why it’s measured on public test sets with a common metric, the WER.
Echo runs speech recognition on its own servers for transcription, live transcription and dictating questions; the audio isn’t sent to third-party recognition services.
