Enhancing feature compression and reconstruction in time-series and image domains with pseudo-overlap and pseudo-grouping pooling functions

Deep learning techniques are widely used for compressing and reconstructing images and time-series features, thanks to their ability to learn efficient latent representations. In this work, we introduce two families of pooling functions -pseudo-overlap and pseudo-grouping- and integrate them within...

ver descrição completa

Detalhes bibliográficos
Autores: Carollo, Matteo, Ferrero Jaurrieta, Mikel, Paiva, Rui, Bardozzo, Francesco, Tagliaferri, Roberto
Formato: artículo
Estado:Versión publicada
Fecha de publicación:2026
País:España
Recursos:Universidad Pública de Navarra
Repositorio:Academica-e. Repositorio Institucional de la Universidad Pública de Navarra
OAI Identifier:oai:dnet:academicae__::bdb7ee9e6791e78d598e56acfb2043e5
Acesso em linha:https://hdl.handle.net/2454/56702
Access Level:acceso abierto
Palavra-chave:Aggregation functions
Compression and reconstruction
Deep learning
Feature extraction
Pooling
id ES_7a23f22b2ef43633446d1c3b7ca7fbab
oai_identifier_str oai:dnet:academicae__::bdb7ee9e6791e78d598e56acfb2043e5
network_acronym_str ES
network_name_str España
repository_id_str
spelling Enhancing feature compression and reconstruction in time-series and image domains with pseudo-overlap and pseudo-grouping pooling functionsCarollo, MatteoFerrero Jaurrieta, MikelPaiva, RuiBardozzo, FrancescoTagliaferri, RobertoAggregation functionsCompression and reconstructionDeep learningFeature extractionPoolingDeep learning techniques are widely used for compressing and reconstructing images and time-series features, thanks to their ability to learn efficient latent representations. In this work, we introduce two families of pooling functions -pseudo-overlap and pseudo-grouping- and integrate them within an autoencoder to improve feature extraction and latent space organization. Unlike conventional pooling methods, these functions adaptively modulate feature aggregation, allowing better preservation of structural information during compression. We evaluate the proposed pooling mechanisms on both image (MNIST, FashionMNIST, CIFAR10, SVHN, CIFAR100, ImageNet) and time-series datasets (Vasicek, J-Vasicek, GMB, J-GMB, MRD, J-MRD), demonstrating enhanced reconstruction accuracy. Comparative experiments show that pseudo-overlap and pseudo-grouping outperform traditional pooling layers, especially in datasets with complex structures, highlighting the importance of carefully designing pooling operations for optimal performance.This work was partly supported by the project ’Future Artificial Intelligence Research (FAIR)’ Project Code PE00000013, Spoke 3 ’Resilient AI’ under the National Recovery and Resilience Plan (PNRR), Mission 4 ’Education and Research’ Component 2 ’From Research to Business’ Investment 1.3, funded by the European Union - NEXTGENERATIONEU - RAISE - WP2 (CUP E63C22002150007 ). The authors wish to thank the CINECA Super Computing Application and Innovation department (SCAI) of Bologna (Italy) for granting MARCONI100.ElsevierEstadística, Informática y MatemáticasEstatistika, Informatika eta Matematika2026info:eu-repo/semantics/articleinfo:eu-repo/semantics/publishedVersionapplication/pdfhttps://hdl.handle.net/2454/56702reponame:Academica-e. Repositorio Institucional de la Universidad Pública de Navarrainstname:Universidad Pública de NavarraInglés© 2026 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license.https://creativecommons.org/licenses/by-nc-nd/4.0/info:eu-repo/semantics/openAccessoai:dnet:academicae__::bdb7ee9e6791e78d598e56acfb2043e52026-06-17T12:41:47Z
dc.title.none.fl_str_mv Enhancing feature compression and reconstruction in time-series and image domains with pseudo-overlap and pseudo-grouping pooling functions
title Enhancing feature compression and reconstruction in time-series and image domains with pseudo-overlap and pseudo-grouping pooling functions
spellingShingle Enhancing feature compression and reconstruction in time-series and image domains with pseudo-overlap and pseudo-grouping pooling functions
Carollo, Matteo
Aggregation functions
Compression and reconstruction
Deep learning
Feature extraction
Pooling
title_short Enhancing feature compression and reconstruction in time-series and image domains with pseudo-overlap and pseudo-grouping pooling functions
title_full Enhancing feature compression and reconstruction in time-series and image domains with pseudo-overlap and pseudo-grouping pooling functions
title_fullStr Enhancing feature compression and reconstruction in time-series and image domains with pseudo-overlap and pseudo-grouping pooling functions
title_full_unstemmed Enhancing feature compression and reconstruction in time-series and image domains with pseudo-overlap and pseudo-grouping pooling functions
title_sort Enhancing feature compression and reconstruction in time-series and image domains with pseudo-overlap and pseudo-grouping pooling functions
dc.creator.none.fl_str_mv Carollo, Matteo
Ferrero Jaurrieta, Mikel
Paiva, Rui
Bardozzo, Francesco
Tagliaferri, Roberto
author Carollo, Matteo
author_facet Carollo, Matteo
Ferrero Jaurrieta, Mikel
Paiva, Rui
Bardozzo, Francesco
Tagliaferri, Roberto
author_role author
author2 Ferrero Jaurrieta, Mikel
Paiva, Rui
Bardozzo, Francesco
Tagliaferri, Roberto
author2_role author
author
author
author
dc.contributor.none.fl_str_mv Estadística, Informática y Matemáticas
Estatistika, Informatika eta Matematika
dc.subject.none.fl_str_mv Aggregation functions
Compression and reconstruction
Deep learning
Feature extraction
Pooling
topic Aggregation functions
Compression and reconstruction
Deep learning
Feature extraction
Pooling
description Deep learning techniques are widely used for compressing and reconstructing images and time-series features, thanks to their ability to learn efficient latent representations. In this work, we introduce two families of pooling functions -pseudo-overlap and pseudo-grouping- and integrate them within an autoencoder to improve feature extraction and latent space organization. Unlike conventional pooling methods, these functions adaptively modulate feature aggregation, allowing better preservation of structural information during compression. We evaluate the proposed pooling mechanisms on both image (MNIST, FashionMNIST, CIFAR10, SVHN, CIFAR100, ImageNet) and time-series datasets (Vasicek, J-Vasicek, GMB, J-GMB, MRD, J-MRD), demonstrating enhanced reconstruction accuracy. Comparative experiments show that pseudo-overlap and pseudo-grouping outperform traditional pooling layers, especially in datasets with complex structures, highlighting the importance of carefully designing pooling operations for optimal performance.
publishDate 2026
dc.date.none.fl_str_mv 2026
dc.type.none.fl_str_mv info:eu-repo/semantics/article
info:eu-repo/semantics/publishedVersion
format article
status_str publishedVersion
dc.identifier.none.fl_str_mv https://hdl.handle.net/2454/56702
url https://hdl.handle.net/2454/56702
dc.language.none.fl_str_mv Inglés
language_invalid_str_mv Inglés
dc.rights.none.fl_str_mv https://creativecommons.org/licenses/by-nc-nd/4.0/
info:eu-repo/semantics/openAccess
rights_invalid_str_mv https://creativecommons.org/licenses/by-nc-nd/4.0/
eu_rights_str_mv openAccess
dc.format.none.fl_str_mv application/pdf
dc.publisher.none.fl_str_mv Elsevier
publisher.none.fl_str_mv Elsevier
dc.source.none.fl_str_mv reponame:Academica-e. Repositorio Institucional de la Universidad Pública de Navarra
instname:Universidad Pública de Navarra
instname_str Universidad Pública de Navarra
reponame_str Academica-e. Repositorio Institucional de la Universidad Pública de Navarra
collection Academica-e. Repositorio Institucional de la Universidad Pública de Navarra
repository.name.fl_str_mv
repository.mail.fl_str_mv
_version_ 1869411415663575040
score 15,812429

Registos relacionados