Título:
|
Randomly weighted CNNs for (music) audio classification
|
Autor/a:
|
Pons Puig, Jordi; Serra, Xavier
|
Abstract:
|
Comunicació presentada a: ICASSP 2019 - 2019 IEEE International Conference on Acoustics, Speech and Signal Processing celebrat de 12 al 17 de maig de 2019 a Brighton, Regne Unit. |
Abstract:
|
The computer vision literature shows that randomly weighted neural networks perform reasonably as feature extractors. Following this idea, we study how non-trained (randomly weighted) convolutional neural networks perform as feature extractors for (music) audio classification tasks. We use features extracted from the embeddings of deep architectures as input to a classifier - with the goal to compare classification accuracies when using different randomly weighted architectures. By following this methodology, we run a comprehensive evaluation of the current architectures for audio classification, and provide evidence that the architectures alone are an important piece for resolving (music) audio problems using deep neural networks. |
Abstract:
|
This work is supported by the Maria de Maeztu Programme (MDM-2015-0502), and we are grateful for the GPUs donated by NVidia. |
Materia(s):
|
-Random -Neural networks -Audio -ELM -SVM |
Derechos:
|
© Jordi Pons, Xavier Serra. Licensed under a Creative
Commons Attribution 4.0 International License (CC BY 4.0). Attribution:
Jordi Pons, Xavier Serra. “Randomly weighted CNNs for (music)
audio classification
https://creativecommons.org/licenses/by/4.0/
|
Tipo de documento:
|
Objeto de conferencia Artículo - Versión publicada |
Editor:
|
Institute of Electrical and Electronics Engineers (IEEE)
|
Compartir:
|
|