This chapter draws on material from the 2012 paper “The Intervalgram: An Audio Feature for Large-scale Melody Recognition” by Thomas C. Walters, David A. Ross, and Richard F. Lyon (Walters et al., 2013).

In this chapter, we review a system for representing the melodic content of short pieces of audio using a novel chroma-based representation known as the “intervalgram,” which is a summary of the local pattern of musical intervals in a segment of music. We introduced chroma as pitch within an octave in Section 4.7. The intervalgram is based on a chroma representation derived from the pitchogram, or temporal profile of the stabilized auditory image. Each intervalgram frame is made locally key invariant by means of a “soft” pitch transposition to a local reference. Intervalgrams are generated for a piece of music using multiple overlapping windows. These sets of intervalgrams are used as the basis of a system for detection of identical melodies across a database of music. Using a dynamic-programming-like approach for comparisons between a reference and the song database, performance was evaluated on the dataset. A first test of an intervalgram-based system on this dataset yields a precision at top-1 of 53.8%, with a precision–recall curve that shows very high precision up to moderate recall, suggesting that the intervalgram is adept at identifying the easier-to-match cover songs in the dataset with high robustness. The intervalgram is designed to support locality-sensitive hashing, such that an index lookup from each single intervalgram feature has a moderate probability of retrieving a match, with relatively few false matches.