Cantonese Voice Benchmark粵語語音基準

Questions

Straight answers.

Do I need a full voice stack?
No. ASR and LLM are the baseline and give you a Core Score. TTS and end-to-end are optional and open the Full Score. The leaderboard shows both, so a strong recogniser is never hidden behind a missing speech synthesiser.
How long does integration take?
Two to three hours for ASR, about a day for the full stack. You implement a WebSocket server and we connect to it; there is a scaffold and a conformance checker so you can get to green locally before exposing anything.
Why do you call my endpoint instead of taking a file?
Because latency and turn-taking cannot be measured from a file, and those are 25% of the score. A batch upload path exists as a fallback, but it is Core Score only and labelled on the leaderboard.
What happens if my endpoint goes down mid-window?
The run reports insufficient data rather than a bad score, and it does not consume your quota. Failures we caused are attributed to us and re-run for free. A downed endpoint is never published as a 100% error rate, because that would be a false statement about your system.
Can I see which clips I got wrong?
On practice runs, yes, with the reference transcript alongside your output. On scored runs you see per-item status and your own response, but not the ground truth.
Who judges the LLM pillar?
Two model families from different providers, three runs each, median taken. A judge scoring its own vendor is a conflict, so agreement between families is published. Before the first scored window the judge is validated against human experts and the correlation is published whatever it says.
Who rates the speech?
Paid native Cantonese listeners who pass a calibration set first. They never see which system produced a sample, playback order is randomised, and ratings are only accepted once the audio has actually finished playing.
Will I be named?
Only if you choose to be. Codename by default, opt-in to be named, reversible until publication.
Can I train on the data?
No. You may use it to produce submissions during the edition, and copies are deleted within thirty days of it closing.
Why are three teams sharing rank 2?
Because their confidence intervals overlap. With thirty clips a difference of a fraction of a percent is noise, and printing a strict order over noise is false precision. Tied means we cannot tell them apart, and saying so is more useful than inventing a winner.

Something not answered here? Get in touch.