Tue, 11/13/2012 - 5:30pm - 6:50pm
CCRMA Classroom [Knoll 217]
A method to both identify and synchronize multiple videos of the same event within a large collections of videos and/or audio files will be discussed. Landmark-based audio fingerprinting is used to match multiple recordings of the same event together and time-synchronize each file within the groups. Compared to prior work, we offer improvements towards event identification and a new synchronization refinement method that resolves inconsistent estimates and allows non-overlapping content to be synchronized within larger groups of recordings. Furthermore, the audio fingerprinting-based synchronization/clustering method will be presented as a time-difference-of-arrival estimation method via cross-correlation performed on a non-linearly transformed signal.
Nick Bryan is a CCRMA PhD student