A community-radio call-in booth can score Riva ASR by counting wrong words or by asking whether the producer still routed the caller to the right topic bin. Which measure matches the booth’s real job?
Select an answer to reveal the explanation.
Short Explanation
If the booth’s job is getting callers into the right topic bin, celebrate that routing win. Perfect spelling on every word is nice, but it is not the job when intent survives.
Full Explanation
Metrics should match the decision. When the booth’s purpose is topic routing, downstream intent success is the fitting ASR experiment measure. Verbatim word error matters for transcript archives, but CLIP-style image similarity and GPU reservation counts do not reflect the producer’s routing task.