Tonal description of polyphonic audio for music content processing

We present a method to extract a description of the tonal aspects of music from polyphonic audio signals. We define this tonal description using different levels of abstraction, differentiating between low-level signal descriptors and high-level textual labels. We also establish different temporal scales for description, defining some features as being attached to a certain time instant, and other global descriptors as related to a wider segment. The description is validated by estimating the key of a piece. We also propose the description as a tonal representation of the polyphonic audio signal to measure tonal similarity between audio excerpts and to establish the tonal structure of a musical piece.