Sequential Modeling of Temporal Timbre Series for Popular Music Sub-Genre Analyses Using Deep Bidirectional Encoder Representations from Transformers

Geng S, Ren G, Pan X, Zysman J, Ogihara M. Sequential Modeling of Temporal Timbre Series for Popular Music Sub-Genre Analyses Using Deep Bidirectional Encoder Representations from Transformers. In: AES Conference: 2020 AES International Conference on Audio for Virtual and Augmented Reality (August 2020). Audio Engineering Society; 2020. Paper 10470. Available from: https://aes.org/publications/elibrary-page/?id=21147

@inproceedings{Geng2020_21147,
  author = {Geng, Shijia and Ren, Gang and Pan, Xu and Zysman, Joel and Ogihara, Mitsu},
  title = {{Sequential Modeling of Temporal Timbre Series for Popular Music Sub-Genre Analyses Using Deep Bidirectional Encoder Representations from Transformers}},
  booktitle = {AES Conference: 2020 AES International Conference on Audio for Virtual and Augmented Reality (August 2020)},
  note = {Paper 10470},
  year = {2020},
  month = aug,
  publisher = {Audio Engineering Society},
  url = {https://aes.org/publications/elibrary-page/?id=21147}
}

TY  - CPAPER
TI  - Sequential Modeling of Temporal Timbre Series for Popular Music Sub-Genre Analyses Using Deep Bidirectional Encoder Representations from Transformers
AU  - Geng, Shijia
AU  - Ren, Gang
AU  - Pan, Xu
AU  - Zysman, Joel
AU  - Ogihara, Mitsu
T2  - AES Conference: 2020 AES International Conference on Audio for Virtual and Augmented Reality (August 2020)
M1  - Paper 10470
PY  - 2020
DA  - 2020/08/06
UR  - https://aes.org/publications/elibrary-page/?id=21147
PB  - Audio Engineering Society
LA  - en
AB  - The timbral analysis from spectrographic features of popular music sub-genres (or micro-genres) presents unique challenges to the field of the computational auditory scene analysis, which is caused by the adjacencies among sub-genres and the complex sonic scenes from sophisticated musical textures and production processes. This paper presents a timbral modeling tool based on a modified deep learning natural language processing model. It treats the time frames in spectrograms as words in natural languages to explore the temporal dependencies. The modeling performance metrics obtained from the fine-tuned classifier of the modified Deep Bidirectional Encoder Representations from Transformers (BERT) model show strong semantic modeling performances with different temporal settings. Designed as an automatic feature engineering tool, the proposed framework provides a unique solution to the semantic modeling and representation tasks for objectively understanding of subtle musical timbral patterns from highly similar musical genres.
ER  -