Designing efficient architectures for modeling temporal features with convolutional neural networks

Mostra el registre complet Registre parcial de l'ítem

• dc.contributor.author Pons Puig, Jordica
• dc.contributor.author Serra, Xavierca
• dc.date.accessioned 2018-02-06T10:37:10Z
• dc.date.available 2018-02-06T10:37:10Z
• dc.date.issued 2017
• dc.description Comunicació presentada a la 2017 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), celebrada els dies 5 a 9 de març de 2017 a Nova Orleans, Louisiana (EUA).
• dc.description.abstract Many researchers use convolutional neural networks with small rectangular filters for music (spectrograms) classification. First, we discuss why there is no reason to use this filters setup by default and second, we point that more efficient architectures could be implemented if the characteristics of the music features are considered during the design process. Specifically, we propose a novel design strategy that might promote more expressive and intuitive deep learning architectures by efficiently exploiting the representational capacity of the first layer - using different filter shapes adapted to fit musical concepts within the first layer. The proposed architectures are assessed by measuring their accuracy in predicting the classes of the Ballroom dataset. We also make available1 the used code (together with the audio-data) so that this research is fully reproducible.en
• dc.description.sponsorship This work is partially supported by the Maria de Maeztu Units of Excellence Programme (MDM-2015-0502).
• dc.format.mimetype application/pdf
• dc.identifier.citation Pons J, Serra X. Designing efficient architectures for modeling temporal features with convolutional neural networks. In: 2017 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP); 2017 Mar 5-9; New Orleans, LA. Piscataway (NJ): IEEE; 2017. p. 2472-6. DOI: 10.1109/ICASSP.2017.7952601
• dc.identifier.doi http://dx.doi.org/10.1109/ICASSP.2017.7952601
• dc.identifier.uri http://hdl.handle.net/10230/33811
• dc.language.iso eng
• dc.publisher Institute of Electrical and Electronics Engineers (IEEE)ca
• dc.relation.ispartof 2017 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP); 2017 Mar 5-9; New Orleans, LA. Piscataway (NJ): IEEE; 2017. p. 2472-6.
• dc.relation.isreferencedby https://github.com/jordipons/ICASSP2017
• dc.rights © 2017 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works. The final published article can be found at http://ieeexplore.ieee.org/document/7952601/
• dc.rights.accessRights info:eu-repo/semantics/openAccess
• dc.subject.keyword Convolutional neural networksen
• dc.subject.keyword Deep learningen
• dc.subject.keyword Musicen
• dc.subject.keyword Classificationen
• dc.subject.keyword Information retrievalen
• dc.title Designing efficient architectures for modeling temporal features with convolutional neural networksca
• dc.type info:eu-repo/semantics/conferenceObject
• dc.type.version info:eu-repo/semantics/acceptedVersion