Classification of one-dimensional non-stationary signals using the Wigner-Ville distribution in convolutional neural networks
(2017) 25th European Signal Processing Conference, EUSIPCO 2017 p.326-330- Abstract
In this paper we argue that the Wigner-Ville distribution (WVD), instead of the spectrogram, should be used as basic input into convolutional neural network (CNN) based classification schemes. The WVD has superior resolution and localization as compared to other time-frequency representations. We present a method where a large-size kernel may be learned from the data, to enhance features important for classification. We back up our claims with theory, as well as application on simulated examples and show superior performance as compared to the commonly used spectrogram.
Please use this url to cite or link to this publication:
https://lup.lub.lu.se/record/e6f085a6-088f-493d-a25b-6eeebf89ab2c
- author
- Brynolfsson, Johan LU and Sandsten, Maria LU
- organization
- publishing date
- 2017-10-23
- type
- Chapter in Book/Report/Conference proceeding
- publication status
- published
- subject
- host publication
- 25th European Signal Processing Conference, EUSIPCO 2017
- article number
- 8081222
- pages
- 5 pages
- publisher
- IEEE - Institute of Electrical and Electronics Engineers Inc.
- conference name
- 25th European Signal Processing Conference, EUSIPCO 2017
- conference location
- Kos, Greece
- conference dates
- 2017-08-28 - 2017-09-02
- external identifiers
-
- scopus:85041504101
- ISBN
- 9780992862671
- DOI
- 10.23919/EUSIPCO.2017.8081222
- language
- English
- LU publication?
- yes
- id
- e6f085a6-088f-493d-a25b-6eeebf89ab2c
- date added to LUP
- 2018-02-22 11:49:26
- date last changed
- 2022-04-25 05:49:27
@inproceedings{e6f085a6-088f-493d-a25b-6eeebf89ab2c, abstract = {{<p>In this paper we argue that the Wigner-Ville distribution (WVD), instead of the spectrogram, should be used as basic input into convolutional neural network (CNN) based classification schemes. The WVD has superior resolution and localization as compared to other time-frequency representations. We present a method where a large-size kernel may be learned from the data, to enhance features important for classification. We back up our claims with theory, as well as application on simulated examples and show superior performance as compared to the commonly used spectrogram.</p>}}, author = {{Brynolfsson, Johan and Sandsten, Maria}}, booktitle = {{25th European Signal Processing Conference, EUSIPCO 2017}}, isbn = {{9780992862671}}, language = {{eng}}, month = {{10}}, pages = {{326--330}}, publisher = {{IEEE - Institute of Electrical and Electronics Engineers Inc.}}, title = {{Classification of one-dimensional non-stationary signals using the Wigner-Ville distribution in convolutional neural networks}}, url = {{http://dx.doi.org/10.23919/EUSIPCO.2017.8081222}}, doi = {{10.23919/EUSIPCO.2017.8081222}}, year = {{2017}}, }