숟가락 그만 얹어
PerformanceRNN 본문
Preprocess
Event-based representation: note-on, note-off, time-shift, velocity에 대한 event를 one-hot vector로 표현하여 sequentially 나열한다.
Model
- Autoregressive LSTM
- Input as 413-dimensional one-hot vectors
- Teacher forcing for training
- Sampling output from beam search
이외에도 music generation에 대한 전반적인 개요를 설명해주는 좋은 논문.
References
[1] S. Oore et al., This Time with Feeling: Learning Expressive Musical Performance, Neural Computing and Applications 2018
'Research > Music Generation' 카테고리의 다른 글
Symbolic Music Motif Generation (0) | 2023.02.10 |
---|---|
MusicVAE (0) | 2021.03.12 |