Acoustic pattern identification using spectral characteristics to synchronize audio and/or video

摘要

Embodiments of the invention relate generally to computing devices and systems, software, computer programs, applications, and user interfaces for identifying acoustic patterns, and more particularly, to determining equivalent portions of audio using spectral characteristics to, for example, synchronize audio and/or video captured at multiple cameras or different intervals of time.

主权项

1. A computer-readable storage medium, excluding signals per se, comprising executable instructions to perform:
analyzing a plurality of content items, the content items including a video portion and an audio portion, and the content items being from multiple takes captured from a same scene in substantially a consecutive, non-overlapping fashion during different intervals of time; determining matching audio portions of the content items based at least in part on spectral signatures of the audio portions of the content items, the spectral signatures differing from each other due to deviations comprising different rates of delivery of speech or differing utterances in the audio portions of the content items; said determining comprises calculating that the spectral signatures of the matched audio portions are substantially acoustically equivalent regardless of the deviations comprising different rates of delivery of speech or differing utterances; and synchronizing the video portions of the content items based at least in part on said determining matching audio portions of the content items.