3 datasets found

Tags: conversational speech

Filter Results

  • CallHome

    The dataset used for comparing human and machine transcription errors in conversational speech.

  • Buckeye Speech Corpus

    The English dataset consists of approximately 300,000 words spoken by 40 speakers from Central Ohio in conversational settings with an interviewer.

  • Switchboard

    Human speech data comprises a rich set of domain factors such as accent, syntactic and semantic variety, or acoustic environment.
You can also access this registry using the API (see API Docs).