Aligned Hierarchies: A Multi-Scale Structure-Based Representation for Music-Based Data Streams

We introduce aligned hierarchies, a low-dimensional representation for music-based data streams, such as recordings of songs or digitized representations of scores. The aligned hierarchies encode all hierarchical decompositions of repeated elements from a high-dimensional and noisy music-based data stream into one object. These aligned hierarchies can be embedded into a classification space with a natural notion of distance. We construct the aligned hierarchies by finding, encoding, and synthesizing all repeated structure present in a music-based data stream. For a data set of digitized scores, we conducted experiments addressing the fingerprint task that achieved perfect precisionrecall values. These experiments provide an initial proof of concept for the aligned hierarchies addressing MIR tasks.