US 12,260,875 B2
Phrase extraction for ASR models
Ehsan Amid, Mountain View, CA (US); Om Dipakbhai Thakkar, Sunnyvale, CA (US); Rajiv Mathews, Sunnyvale, CA (US); and Francoise Beaufays, Mountain View, CA (US)
Assigned to Google LLC, Mountain View, CA (US)
Filed by Google LLC, Mountain View, CA (US)
Filed on Mar. 19, 2024, as Appl. No. 18/609,362.
Application 18/609,362 is a continuation of application No. 17/643,848, filed on Dec. 13, 2021, granted, now 11,955,134.
Claims priority of provisional application 63/264,836, filed on Dec. 2, 2021.
Prior Publication US 2024/0221772 A1, Jul. 4, 2024
This patent is subject to a terminal disclaimer.
Int. Cl. G10L 21/0332 (2013.01); G10L 15/06 (2013.01); G10L 15/08 (2006.01); G10L 21/10 (2013.01)
CPC G10L 21/0332 (2013.01) [G10L 15/063 (2013.01); G10L 15/08 (2013.01); G10L 21/10 (2013.01)] 18 Claims
OG exemplary drawing
 
1. A computer-implemented method executed on data processing hardware that causes the data processing hardware to perform operations comprising:
obtaining audio data characterizing an utterance and a corresponding ground-truth transcription of the utterance;
processing the ground-truth transcription to identify a particular phrase included in the ground-truth transcription that is associated with sensitive data;
modifying the audio data to obfuscate the particular phrase recited in the utterance;
processing, using a trained automated speech recognition (ASR) model, the modified audio data to generate a predicted transcription of the utterance;
determining whether the predicted transcription includes the particular phrase or another phrase substituted for the particular phrase from the ground-truth transcription that is associated with a same category of information as the particular phrase by comparing the predicted transcription of the utterance to the ground-truth transcription of the utterance; and
when the predicted transcription includes the other phrase substituted for the particular phrase, generating an output indicating that the trained ASR model leaked the other phrase from a training data set used to train the ASR model.