Analysis, Visualization, and Transformation of audio signals using Dictionary-Based Methods

This article provides an overview of dictionary-based methods (DBMs), and reviews recent work in the application of such methods to working with audio and music signals. As Fourier analysis is to additive synthesis, DBMs can be seen as the analytical counterpart to a generalized granular synthesis, where a sound is built by combining heterogeneous atoms selected from a user-defined dictionary. As such, DBMs provide novel ways for analyzing and visualizing audio signals, creating multiresolution descriptions of their contents, and designing sound transformations unique to a description of audio in terms of atoms.