TY - GEN
T1 - Signal reconstruction by means of Embedding, Clustering and AutoEncoder Ensembles
AU - Mio, Corrado
AU - Gianini, Gabriele
N1 - Funding Information:
The authors acknowledge the support of the ICT Fund at EBTIC/Khalifa University, Abu Dhabi, UAE. The work was partially funded by the EU H2020 Research Programme, within the projects Toreador (Grant Agreement No. 688797), Threat-Arrest (Grant Agreement No. 786890) and Concordia (Grant Agreement No. 830927).
Publisher Copyright:
© 2019 IEEE.
PY - 2019/6
Y1 - 2019/6
N2 - We study the denoising and reconstruction of corrupted signals by means of AutoEncoder ensembles. In order to guarantee experts' diversity in the ensemble, we apply, prior to learning, a dimensional reduction pass (to map the examples into a suitable Euclidean space) and a partitional clustering pass: each cluster is then used to train a distinct AutoEncoder. We study the approach with an audio file benchmark: the original signals are artificially corrupted by Doppler effect and reverb. The results support the comparative effectiveness of the approach, w.r.t. the approach based on a single AutoEncoder. The processing pipeline using Local Linear Embedding, k means, then k Convolutional Denoising AutoEncoders reduces the reconstruction error by 35% w.r.t. the baseline approach.
AB - We study the denoising and reconstruction of corrupted signals by means of AutoEncoder ensembles. In order to guarantee experts' diversity in the ensemble, we apply, prior to learning, a dimensional reduction pass (to map the examples into a suitable Euclidean space) and a partitional clustering pass: each cluster is then used to train a distinct AutoEncoder. We study the approach with an audio file benchmark: the original signals are artificially corrupted by Doppler effect and reverb. The results support the comparative effectiveness of the approach, w.r.t. the approach based on a single AutoEncoder. The processing pipeline using Local Linear Embedding, k means, then k Convolutional Denoising AutoEncoders reduces the reconstruction error by 35% w.r.t. the baseline approach.
UR - http://www.scopus.com/inward/record.url?scp=85078881816&partnerID=8YFLogxK
U2 - 10.1109/ISCC47284.2019.8969655
DO - 10.1109/ISCC47284.2019.8969655
M3 - Conference contribution
AN - SCOPUS:85078881816
T3 - Proceedings - IEEE Symposium on Computers and Communications
BT - Proceedings - IEEE Symposium on Computers and Communications
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 2019 IEEE Symposium on Computers and Communications, ISCC 2019
Y2 - 29 June 2019 through 3 July 2019
ER -