Extraction and matching of characteristic fingerprints from audio signals

摘要

An audio fingerprint is extracted from an audio sample, where the fingerprint contains information that is characteristic of the content in the sample. The fingerprint may be generated by computing an energy spectrum for the audio sample, resampling the energy spectrum logarithmically in the time dimension, transforming the resampled energy spectrum to produce a series of feature vectors, and computing the fingerprint using differential coding of the feature vectors. The generated fingerprint can be compared to a set of reference fingerprints in a database to identify the original audio content.

主权项

1. A method comprising:
determining, by a computing device, a test audio fingerprint; computing, by the computing device, an index value for the test audio fingerprint; using the index value, determining, by the computing device, a group of candidate audio fingerprints; computing, by the computing device, a bit error rate between the test audio fingerprint and each candidate audio fingerprint in the group; and determining, by the computing device, that the test audio fingerprint matches a candidate audio fingerprint based on the bit error rate.