Speech Recognition
FoundationsApplications and Capabilities
Also called: Automatic Speech Recognition, ASR
Speech recognition is the technology that converts spoken audio into machine-readable text. Also known as automatic speech recognition, or ASR, it is built on deep neural networks and underpins enterprise voice assistants, contact centre transcription, medical dictation and real-time meeting translation across languages.
In practice
Accuracy is not a single number: it collapses on accents, industry vocabulary and people talking over each other, and the vendor benchmark was measured on clean audio. Test on your own recordings before committing, because in a contact centre a few points of word error rate is the difference between usable transcripts and a second manual pass.