Adversarial signal denoising with encoder-decoder networks

Casas L, Klimmek A, Navab N, Belagiannis V (2021)


Publication Type: Conference contribution

Publication year: 2021

Publisher: European Signal Processing Conference, EUSIPCO

Book Volume: 2021-January

Pages Range: 1467-1471

Conference Proceedings Title: European Signal Processing Conference

Event location: Amsterdam, NLD

ISBN: 9789082797053

DOI: 10.23919/Eusipco47968.2020.9287738

Abstract

The presence of noise is common in signal processing regardless the signal type. Deep neural networks have shown good performance in noise removal, especially on the image domain. In this work, we consider deep neural networks as a denoising tool where our focus is on one dimensional signals. We introduce an encoder-decoder architecture to denoise signals, represented by a sequence of measurements. Instead of relying only on the standard reconstruction error to train the encoder-decoder network, we treat the task of denoising as distribution alignment between the clean and noisy signals. Then, we propose an adversarial learning formulation where the goal is to align the clean and noisy signal latent representation given that both signals pass through the encoder. In our approach, the discriminator has the role of detecting whether the latent representation comes from clean or noisy signals. We evaluate on electrocardiogram and motion signal denoising; and show better performance than learning-based and non-learning approaches.

Authors with CRIS profile

Involved external institutions

How to cite

APA:

Casas, L., Klimmek, A., Navab, N., & Belagiannis, V. (2021). Adversarial signal denoising with encoder-decoder networks. In European Signal Processing Conference (pp. 1467-1471). Amsterdam, NLD: European Signal Processing Conference, EUSIPCO.

MLA:

Casas, Leslie, et al. "Adversarial signal denoising with encoder-decoder networks." Proceedings of the 28th European Signal Processing Conference, EUSIPCO 2020, Amsterdam, NLD European Signal Processing Conference, EUSIPCO, 2021. 1467-1471.

BibTeX: Download