SWITCHBOARD¶
Godfrey, J. J., Holliman, E. C., & McDaniel, J. (1992). SWITCHBOARD: Telephone Speech Corpus for Research and Development. IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP), 1, 517-520.
Cited by¶
1 citation across 1 artifact.
Each citation links to the sentence it supports in the citing article.
Primes¶
- Training Serving Skew
- Lab-to-clinic translation: a procedure that performs in controlled trials underperforming in ordinary practice as efficacy and effectiveness diverge for systematic reasons. Tutorial-to-production code: patterns learned from clean tutorial examples breaking against messy data, concurrency, and partial failures. Speech recognition trained on read speech and deployed against spontaneous speech — a classic instance with the same structural shape.
This sourceThe benchmark corpus of spontaneous conversational telephone speech; recognition error rates on it run far above those on read speech, the canonical read-vs-spontaneous ASR gap.
Supported in partVerified against the publisher's abstract
Abstract supplies the spontaneous conversational-telephone-speech corpus, but not the claim that recognizers trained on read speech degrade against it; that gap is unstated here.
“SWITCHBOARD is a large multispeaker corpus of conversational speech and text which should be of interest to researchers in speaker authentication and large vocabulary speech recognition. About 2500 conversations by 500 speakers from around the US were collected automatically over T1 lines at Texas Instruments.”
- Lab-to-clinic translation: a procedure that performs in controlled trials underperforming in ordinary practice as efficacy and effectiveness diverge for systematic reasons. Tutorial-to-production code: patterns learned from clean tutorial examples breaking against messy data, concurrency, and partial failures. Speech recognition trained on read speech and deployed against spontaneous speech — a classic instance with the same structural shape.
Verification¶
Does it exist? Confirmed. This work's DOI resolves to a registered record, which fixes its identity. That is all it fixes.
Does it back the claim? Read against the text for 1 of 1 citation: 1 supported in part. Each verdict is shown under its citation below, with what in the work backs the sentence.
Support is checked per citation rather than per work — the same source can be cited soundly in one article and wrongly in another. Per-citation recording began recently, so a citation with no recorded check is a gap in the record rather than evidence it went unchecked.
See how references were verified.
Registry ID ref:1b68d71900d2 · see in the full table