GigaSpeechBench — Medical
The Chinese and English medical subsets support recognition evaluation on domain-specific speech and terminology.
Visit official source (opens in a new tab)Research use
Compare context prompting and hotword handling using the project’s medical-domain evaluation and reference transcripts.
Limitations
The subsets evaluate domain vocabulary. Speaker ages and consultation-specific sampling are unconfirmed.
License and access
- Dataset license
- CC BY 4.0 (opens in a new tab)
- Access status
- Direct access
Load the MED-CH and MED-EN test configurations from the publisher’s public WebDataset release.