Speaker: Prof. Dasaem Jeong
Title: The Challenges in Deep Learning on Symbolic Music
Time: Mon, Apr 24 2023, 14:30–15:45 
Place: N25 3229, Paik Nam June Hall  
(You can also join online: https://kaist.zoom.us/j/81348247967?pwd=WWJseG5jR0ZhSmxiSVlLN0VqMjc5UT09)

Bio:
2021-Present: Assistant Professor @ Dept. of Art & Technology, Sogang University
2020-2021: Research Scientist @ T-Brain X, SK telecom
2020: Ph.D at GSCT, KAIST

Abstract: 
Music data can exist in the form of audio implemented as sound, but it can also exist in the form of symbolic music, such as music score or MIDI. This lecture explains how research using deep learning on symbolic music is different from research on music audio, and the challenges involved in handling symbolic music. The development process of a model that was able to generate a perfect score tune in the 2022 AI Music Generation Challenge and other applications of neural networks in traditional Korean music composition and piano performance modeling will be introduced.

(This talk is delivered in English.)