Data-Free Dynamic Compression of CNNs for Tractable Efficiency

Lukas Meiner, Lukas Meiner, Jens Mehnert, Alexandru Condurache, Alexandru Condurache

2025

Abstract

To reduce the computational cost of convolutional neural networks (CNNs) on resource-constrained devices, structured pruning approaches have shown promise in lowering floating-point operations (FLOPs) without substantial drops in accuracy. However, most methods require fine-tuning or specific training procedures to achieve a reasonable trade-off between retained accuracy and reduction in FLOPs, adding computational overhead and requiring training data to be available. To this end, we propose HASTE (Hashing for Tractable Efficiency), a data-free, plug-and-play convolution module that instantly reduces a network’s test-time inference cost without training or fine-tuning. Our approach utilizes locality-sensitive hashing (LSH) to detect redundancies in the channel dimension of latent feature maps, compressing similar channels to reduce input and filter depth simultaneously, resulting in cheaper convolutions. We demonstrate our approach on the popular vision benchmarks CIFAR-10 and ImageNet, where we achieve a 46.72% reduction in FLOPs with only a 1.25% loss in accuracy by swapping the convolution modules in a ResNet34 on CIFAR-10 for our HASTE module.

Paper Citation

in Harvard Style

Meiner L., Mehnert J. and Condurache A. (2025). Data-Free Dynamic Compression of CNNs for Tractable Efficiency. In Proceedings of the 20th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications - Volume 2: VISAPP; ISBN 978-989-758-728-3, SciTePress, pages 196-208. DOI: 10.5220/0013301000003912

in Bibtex Style

@conference{visapp25,
author={Lukas Meiner and Jens Mehnert and Alexandru Condurache},
title={Data-Free Dynamic Compression of CNNs for Tractable Efficiency},
booktitle={Proceedings of the 20th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications - Volume 2: VISAPP},
year={2025},
pages={196-208},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0013301000003912},
isbn={978-989-758-728-3},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 20th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications - Volume 2: VISAPP
TI - Data-Free Dynamic Compression of CNNs for Tractable Efficiency
SN - 978-989-758-728-3
AU - Meiner L.
AU - Mehnert J.
AU - Condurache A.
PY - 2025
SP - 196
EP - 208
DO - 10.5220/0013301000003912
PB - SciTePress