Set of dissimilarity matrices of covers
D = COVERS80(N)
This is one of the datasets used by Liem and Hanjalic . It is build from a set of dissimilarity matrices between musical fragments in a study on the detection of covers.
 Cynthia C. S. Liem, Alan Hanjalic: Cover Song Retrieval
datasets, prdisdata, covers_songs, covers_beethoven, covers_beatles,