Transcript re-sync

摘要

In an aspect, in general, method for aligning an audio recording and a transcript includes receiving a transcript including a plurality of terms, each term of the plurality of terms associated with a time location within a different version of the audio recording, forming a plurality of search terms from the terms of the transcript, determining possible time locations of the search terms in the audio recording, determining a correspondence between time locations within the different version of the audio recording associated with the search terms and the possible time locations of the search terms in the audio recording, and aligning the audio recording and the transcript including updating the time location associated with terms of the transcript based on the determined correspondence.

主权项

1. A computer implemented method for aligning two versions of an audio recording and a transcript comprising:
receiving a first version of the two versions of the audio recording and a second version of the two versions of the audio recording, wherein one version of the two versions of the audio recording includes a modification of the other version of the audio recording; receiving a transcript including a plurality of terms, each term of the plurality of terms associated with a time location within the first version of the audio recording; forming via a search term formation module a plurality of search terms from the terms of the transcript; using an alignment module to:
determine possible time locations of the search terms in the second version of the audio recording;determine a correspondence between time locations within the first version of the audio recording associated with the search terms and the possible time locations of the search terms in the second version of the audio recording; andalign the second version of the audio recording and the transcript including updating the time location associated with terms of the transcript based on the determined correspondence, wherein updating the time location comprises, if a first time difference between a time location of a first term and a time location of a second term within the first version of the audio recording is similar to a second time difference between an updated time location of a first term and an updated time location of a second term then updating a time location associated with a term located between the first term and the second term based on a predicted time location of the term.