Cargando…

Dynamic Image Difficulty-Aware DNN Pruning

Deep Neural Networks (DNNs) have achieved impressive performance in various image recognition tasks, but their large model sizes make them challenging to deploy on resource-constrained devices. In this paper, we propose a dynamic DNN pruning approach that takes into account the difficulty of the inc...

Descripción completa

Detalles Bibliográficos
Autores principales: Pentsos, Vasileios, Spantidi, Ourania, Anagnostopoulos, Iraklis
Formato: Online Artículo Texto
Lenguaje:English
Publicado: MDPI 2023
Materias:
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10224338/
https://www.ncbi.nlm.nih.gov/pubmed/37241531
http://dx.doi.org/10.3390/mi14050908
_version_ 1785050153536716800
author Pentsos, Vasileios
Spantidi, Ourania
Anagnostopoulos, Iraklis
author_facet Pentsos, Vasileios
Spantidi, Ourania
Anagnostopoulos, Iraklis
author_sort Pentsos, Vasileios
collection PubMed
description Deep Neural Networks (DNNs) have achieved impressive performance in various image recognition tasks, but their large model sizes make them challenging to deploy on resource-constrained devices. In this paper, we propose a dynamic DNN pruning approach that takes into account the difficulty of the incoming images during inference. To evaluate the effectiveness of our method, we conducted experiments on the ImageNet dataset on several state-of-art DNNs. Our results show that the proposed approach reduces the model size and amount of DNN operations without the need to retrain or fine-tune the pruned model. Overall, our method provides a promising direction for designing efficient frameworks for lightweight DNN models that can adapt to the varying complexity of input images.
format Online
Article
Text
id pubmed-10224338
institution National Center for Biotechnology Information
language English
publishDate 2023
publisher MDPI
record_format MEDLINE/PubMed
spelling pubmed-102243382023-05-28 Dynamic Image Difficulty-Aware DNN Pruning Pentsos, Vasileios Spantidi, Ourania Anagnostopoulos, Iraklis Micromachines (Basel) Article Deep Neural Networks (DNNs) have achieved impressive performance in various image recognition tasks, but their large model sizes make them challenging to deploy on resource-constrained devices. In this paper, we propose a dynamic DNN pruning approach that takes into account the difficulty of the incoming images during inference. To evaluate the effectiveness of our method, we conducted experiments on the ImageNet dataset on several state-of-art DNNs. Our results show that the proposed approach reduces the model size and amount of DNN operations without the need to retrain or fine-tune the pruned model. Overall, our method provides a promising direction for designing efficient frameworks for lightweight DNN models that can adapt to the varying complexity of input images. MDPI 2023-04-23 /pmc/articles/PMC10224338/ /pubmed/37241531 http://dx.doi.org/10.3390/mi14050908 Text en © 2023 by the authors. https://creativecommons.org/licenses/by/4.0/Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
spellingShingle Article
Pentsos, Vasileios
Spantidi, Ourania
Anagnostopoulos, Iraklis
Dynamic Image Difficulty-Aware DNN Pruning
title Dynamic Image Difficulty-Aware DNN Pruning
title_full Dynamic Image Difficulty-Aware DNN Pruning
title_fullStr Dynamic Image Difficulty-Aware DNN Pruning
title_full_unstemmed Dynamic Image Difficulty-Aware DNN Pruning
title_short Dynamic Image Difficulty-Aware DNN Pruning
title_sort dynamic image difficulty-aware dnn pruning
topic Article
url https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10224338/
https://www.ncbi.nlm.nih.gov/pubmed/37241531
http://dx.doi.org/10.3390/mi14050908
work_keys_str_mv AT pentsosvasileios dynamicimagedifficultyawarednnpruning
AT spantidiourania dynamicimagedifficultyawarednnpruning
AT anagnostopoulosiraklis dynamicimagedifficultyawarednnpruning