"Multimedia Fingerprinting for Content Identification: Model Based Decoding Metrics"
Speaker:
Rohit NainiDate and Time:
January 26, 2012 - 10:10am - 10:30amPresentation Abstract:
Hash-based content identification (ID) is an emerging research area. Multimedia content is identified by matching short fingerprints extracted from it (using robust hashing techniques) to a large database. A variety of algorithms have already been developed for constructing robust hashes as well as efficient string matching. In our work, a fairly general class of structured codes is considered, and a statistical model for the resulting fingerprints and their degraded versions (following miscellaneous content distortions) is proposed and validated. The Maximum-Likelihood fingerprint decoder derived from this model is shown to considerably improve upon previous decoders based on the Hamming metric.