An Interactive Automatic Violin Fingering Recommendation Interface
Vincent K.M. Cheung, Tsung-Ping Chen, Li Su · 2021
Successful execution and musical communication require an effective fingering combination for each note the violinist plays. Fingering annotations on the musical score thus serve as important performance reminders for violinists of all abilities, particularly for beginners. However, fingering information is wholly or partially absent from most violin sheet music and existing fingering generation models require extensive preprocessing with low user-accessibility. Here, we showcase a simple graphical user interface (GUI) that automatically annotates violin fingerings on sheet music saved in the widely-adopted MusicXML format and exports the annotated score into a PDF or CSV file. Our fingering recommendation system is based on a deep learning model that further allows customised user inputs and provides three different fingering modes to best suit the unique background, training goals, and abilities of each violinist.