| Presentation | 2023-11-23 [Poster Presentation] A Study of Complexity Reduction for Classification of Musical Instruments Using Element Selection Ryu Kato, Natsuki Ueno, Nobutaka Ono, Ryo Matsuda, Kazunobu Kondo, |
|---|---|
| PDF Download Page |  PDF download Page Link |
| Abstract(in Japanese) | (See Japanese page) |
| Abstract(in English) | In this study, we propose complexity reduction in convolutional-neural-network (CNN)-based music instruments classification by incorporating an element selection method. Classification of musical instruments is a fundamental technic in various applications such as automatic transcription, musical information retrieval, and automatic application of audio effects. Reducing computational costs is desired in some of these applications. In contrast, element selection is a method that extracts specific components from a feature vector to achieve a lower-dimensional representation. Unlike other dimension reduction techniques like Principal Component Analysis, it offers the advantage of not requiring multiplication. In our study, we aim to reduce the computational complexity in both CNN and the dimension reduction process itself by applying element selection to the input features. The selected elements are optimized to minimize the mean reconstruction error. We experimentally evaluated the changes in computation time and estimation accuracy using a dataset of musical instrument sounds. Our proposed method showed lower degradation in estimation performance compared to random element selection. Additionally, we confirmed that the dimension reduction technique using element selection enables instrument sound classification in a shorter computation time. |
| Keyword(in Japanese) | (See Japanese page) |
| Keyword(in English) | convolutional neural network / musical instruments classification / dimensyonality reduction / element selection |
| Paper # | EA2023-37,EMM2023-68 |
| Date of Issue | 2023-11-16 (EA, EMM) |
| Conference Information | |
| Committee | EMM / EA / ASJ-H |
|---|---|
| Conference Date | 2023/11/23(2days) |
| Place (in Japanese) | (See Japanese page) |
| Place (in English) | |
| Topics (in Japanese) | (See Japanese page) |
| Topics (in English) | [Beginners Session] Engineering/Electro Acoustics, Content Processing, Digital Watermarking, Psychological and Physiological Acoustics, and Related Topics |
| Chair | Michiharu Niimi(Kyushu Inst. of Tech.) / Junki Ono(Tokyo Metropolitan Univ.) |
| Vice Chair | Kotaro Sonoda(Nagasaki Univ.) / Hyunho Kang(NIT, Tokyo) / Takanobu Nishiura(RitsumeikanUniv.) / Yoshinobu Kajikawa(Kansai Univ.) |
| Secretary | Kotaro Sonoda(Hiroshima City Univ.) / Hyunho Kang(Osaka Inst. of Tech.) / Takanobu Nishiura(NTT) / Yoshinobu Kajikawa(Univ. of Tokyo) |
| Assistant | Naofumi Aoki(Hokkaido Univ.) / Kazuaki Nakamura(Tokyo Univ. of Science) / Masato Nakayama(OSU) / Kouhei Yatabe(TUAT) |
| Paper Information | |
| Registration To | Technical Committee on Enriched MultiMedia / Technical Committee on Engineering Acoustics / Auditory Research Meeting |
|---|---|
| Language | JPN |
| Title (in Japanese) | (See Japanese page) |
| Sub Title (in Japanese) | (See Japanese page) |
| Title (in English) | [Poster Presentation] A Study of Complexity Reduction for Classification of Musical Instruments Using Element Selection |
| Sub Title (in English) | |
| Keyword(1) | convolutional neural network |
| Keyword(2) | musical instruments classification |
| Keyword(3) | dimensyonality reduction |
| Keyword(4) | element selection |
| 1st Author's Name | Ryu Kato |
| 1st Author's Affiliation | Tokyo Matropolitan University(Tokyo Metropolitan Univ.) |
| 2nd Author's Name | Natsuki Ueno |
| 2nd Author's Affiliation | Tokyo Matropolitan University(Tokyo Metropolitan Univ.) |
| 3rd Author's Name | Nobutaka Ono |
| 3rd Author's Affiliation | Tokyo Matropolitan University(Tokyo Metropolitan Univ.) |
| 4th Author's Name | Ryo Matsuda |
| 4th Author's Affiliation | Yamaha Corporation(Yamaha Corp.) |
| 5th Author's Name | Kazunobu Kondo |
| 5th Author's Affiliation | Yamaha Corporation(Yamaha Corp.) |
| Date | 2023-11-23 |
| Paper # | EA2023-37,EMM2023-68 |
| Volume (vol) | vol.123 |
| Number (no) | EA-278,EMM-279 |
| Page | pp.pp.51-56(EA), pp.51-56(EMM), |
| #Pages | 6 |
| Date of Issue | 2023-11-16 (EA, EMM) |