Automatic Music Production System Employing Probabilistic Expert Systems

Bocko G, Bocko MF, Headlam D, Lundberg J, Ren G. Automatic Music Production System Employing Probabilistic Expert Systems. In: AES Convention 129. Audio Engineering Society; 2010. Paper 8255. Available from: https://aes.org/publications/elibrary-page/?id=15677

@inproceedings{Bocko2010_15677,
  author = {Bocko, Gregory and Bocko, Mark F. and Headlam, Dave and Lundberg, Justin and Ren, Gang},
  title = {{Automatic Music Production System Employing Probabilistic Expert Systems}},
  booktitle = {AES Convention 129},
  note = {Paper 8255},
  year = {2010},
  month = nov,
  publisher = {Audio Engineering Society},
  url = {https://aes.org/publications/elibrary-page/?id=15677}
}

TY  - CPAPER
TI  - Automatic Music Production System Employing Probabilistic Expert Systems
AU  - Bocko, Gregory
AU  - Bocko, Mark F.
AU  - Headlam, Dave
AU  - Lundberg, Justin
AU  - Ren, Gang
T2  - AES Convention 129
M1  - Paper 8255
PY  - 2010
DA  - 2010/11/06
UR  - https://aes.org/publications/elibrary-page/?id=15677
PB  - Audio Engineering Society
LA  - en
AB  - An automatic music production system based on expert audio engineering knowledge is proposed. An expert system based on a probabilistic graphical model is employed to embed professional audio engineering knowledge and infer automatic production decisions based on musical information extracted from audio files. The production pattern, which is represented as probabilistic graphical model, can be ‘learned’ from the operation data of a human audio engineer or manually constructed from domain knowledge. The authors also discuss the real-time implementation of the proposed automatic production system for live mixing application scenarios. Musical event alignment and prediction algorithms are introduced to improve the time synchronization performance of our production model. The authors conclude with performance evaluations and a brief summary.
ER  -