This page serves as a further material site for  and . As such it contains links to binaries, datasets, benchmarks, etc.
If you find a broken or outdated link, please let me know. Thanks.
A comprehensive list of all kinds of MIR-related datasets can be found at www.audiocontentanalysis.org.
The Accuracy1 results reported in the ISMIR 2017 publication  have been erroneously
computed with a 8% instead of 4% tolerance. This does in no way change the main message of the paper,
that Accuracy1 can be improved through the proposed post-processing procedure.
Corrected results are shown below:
 Hendrik Schreiber, Meinard Müller. Exploiting Global Features for Tempo Octave Correction. In Proceedings of the IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP), Florence, Italy, May 2014.
 Hendrik Schreiber, Meinard Müller. A post-processing procedure for improving music tempo estimates using supervised learning. In Proceedings of the 18th International Society for Music Information Retrieval Conference (ISMIR), Suzhou, China, Oct. 2017.