This page served as the complementary material of . The page contains the dataset and annotation for offset detection, and illustration of the signal to detection function with number of methods including spectral flux, energy and pitch attributes of music.
The dataset contains 63 violin solo excerpts with a total of 1,226 notes derived from the YouTube video clips in  and several sound clips from the website CompositionToday.com. The dataset are categorized into 6 different playing techniques as shown below.
Dataset detail inforamtion
|Techniques||# of clips||# of offsets|
please contact mister2dot4[at]gmail.com or li.sowaterking[at]gmail.com
Below are more figure sampled from the dataset catagoried by playing techniques
- Sul tasto
- Sul ponticello
We appreciate P. M. Brossier (YINFFT), S.Boeck and G.Widmer (Superflux), Colin Raffel et al (MIR_EVAL) kindly sharing the source code online.
This work was supported by the Ministry of Science and Technology of Taiwan under the contracts MOST 102-2221- E-001-004-MY3, MOST 104-2221-E-001-029-MY3, and the Academia Sinica Career Development Program.
 Che-Yuan Liang, Li Su, Hsin-Ming Lin and Yi-Hsuan Yang, "MUSICAL OFFSET DETECTION OF PITCHED INSTRUMENTS: THE CASE OF VIOLIN In Proc. of Int. Soc. Music Information Retrieval Conf., 2015"  L. Su, H.-M. Lin, and Y.-H. Yang. Sparse modeling of magnitude and phase-derived spectra for play- ing technique classification. IEEE/ACM Trans. Audio, Speech and Language Proc., 22(12):2122–2132, 2014. [Online] http://mac.citi.sinica.edu.tw/ violin-playing-technique/.