Applying deep learning to estimate violin-bow's gestural patterns

Eight bow-strokes were recorded by musicians following a musical score with a fixed metronome tempo of quarter-notes at 80 bpm. All gestures were established in G, primarily in the major mode except for tremolo (G minor) and col legno (chromatic). On the violin, two octaves from G3 to G5 covers the whole neck and also all four strings. Eight musicians participated in the recording sessions, with expert models constructed using the data from five violinists; the other three participants were reserved as test cases. The recordings are part of a collaboration with the Royal College of Music in London. Please cite the following publication when using the dataset: Dalmazzo D, Waddell G and Ramírez R (2021) Applying Deep Learning Techniques to Estimate Patterns of Musical Gesture. <em>Front. Psychol.</em> 11:575971. doi: 10.3389/fpsyg.2020.575971 The implementation code is available here.