Drum Synthesis and Rhythmic Transformation with Adversarial Autoencoders
- Submitting institution
-
Birmingham City University
- Unit of assessment
- 11 - Computer Science and Informatics
- Output identifier
- 11Z_OP_E2003
- Type
- E - Conference contribution
- DOI
-
10.1145/3394171.3413519
- Title of conference / published proceedings
- MM '20: Proceedings of the 28th ACM International Conference on Multimedia
- First page
- 2427
- Volume
- -
- Issue
- -
- ISSN
- -
- Open access status
- -
- Month of publication
- October
- Year of publication
- 2020
- URL
-
-
- Supplementary information
-
-
- Request cross-referral to
- -
- Output has been delayed by COVID-19
- No
- COVID-19 affected output statement
- -
- Forensic science
- No
- Criminology
- No
- Interdisciplinary
- No
- Number of additional authors
-
-
- Research group(s)
-
-
- Citation count
- -
- Proposed double-weighted
- No
- Reserve for an output with double weighting
- No
- Additional information
- The presented rhythm transformation for audio completely removes the need for discretised event localisation and classification, allowing for continuous transformative effect processing. While this work is quite recent, it has led to additional, related publications (e.g., Adversarial Synthesis of Drum Sounds in Digital Audio Effects 2020) as well as current collaborative efforts with researchers in multiple international institutions (IRCAM, AIST Japan, University of Coimbra, Universitat Pompeu Fabra) and industry (Sony CSL). In addition, this research has led to major speaking engagements in the field of machine learning for audio (e.g., https://www.aes.org/events/2020/learning/program.cfm).
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -