Using Multiple Feature Extraction with Statistical Models to Categorize Music by Genre

Fields B. Using Multiple Feature Extraction with Statistical Models to Categorize Music by Genre. In: AES Convention 122. Audio Engineering Society; 2007. Paper 7015. Available from: https://aes.org/publications/elibrary-page/?id=14000

@inproceedings{Fields2007_14000,
  author = {Fields, Benjamin},
  title = {{Using Multiple Feature Extraction with Statistical Models to Categorize Music by Genre}},
  booktitle = {AES Convention 122},
  note = {Paper 7015},
  year = {2007},
  month = may,
  publisher = {Audio Engineering Society},
  url = {https://aes.org/publications/elibrary-page/?id=14000}
}

TY  - CPAPER
TI  - Using Multiple Feature Extraction with Statistical Models to Categorize Music by Genre
AU  - Fields, Benjamin
T2  - AES Convention 122
M1  - Paper 7015
PY  - 2007
DA  - 2007/05/06
UR  - https://aes.org/publications/elibrary-page/?id=14000
PB  - Audio Engineering Society
LA  - en
AB  - In recent years, large capacity portable personal music players have become widespread in their use and popularity. Coupled with the exponentially increasing processing power of personal computers and embedded devices, the way people consume and listen to music is ever changing. To facilitate the categorization of these personal music libraries, a system is employed using MPEG-7 feature vectors as well as Mel-Frequency Cepstral Coefficients classified through multiple trained Hidden Markov Models and other statistical methods. The output of these models is then compared and a genre choice is made based on which model gives the best fit. Results from these tests are analyzed and ways to improve the performance of a genre sorting system are discussed.
ER  -