Energy-Efficient Deep Learning using Model Compression

DOI

10.2991/978-94-6239-668-5_92How to use a DOI?

Keywords

Model Compression; Quantization; Pruning

Abstract

Model compression techniques improve the energy efficiency of convolutional neural networks (CNNs) without losing accuracy. Using the CIFAR-10 dataset, we trained a baseline CNN and implemented two key compression methods: pruning and post-training quantization. Our pruned model with 50% sparsity achieved 66.68% accuracy while reducing the model size to 1.38 megabytes (MB). In comparison, the quantized model maintained 67.61% accuracy while being significantly smaller at 1.18 MB. These results demonstrate that quantization not only preserves classification performance but also produces a more compact model suitable for edge deployment and low-power devices. This indicates that quantization can not only decrease the memory footprint of a model but also potentially improve its accuracy. We illustrate the trade-offs between model size and accuracy and highlight quantization as an effective approach for green computing in deep learning. The results support sustainable AI by offering lightweight, energy-efficient models without requiring complex retraining.

Copyright

© 2026 The Author(s)

Open Access

Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

TY  - CONF
AU  - Ayla Kayabaş
PY  - 2026
DA  - 2026/05/14
TI  - Energy-Efficient Deep Learning using Model Compression
BT  - Proceedings of the International Conference on Current Problems in Engineering and Applied Sciences (ICCPEAS 2025)
PB  - Atlantis Press
SP  - 877
EP  - 882
SN  - 2352-5401
UR  - https://doi.org/10.2991/978-94-6239-668-5_92
DO  - 10.2991/978-94-6239-668-5_92
ID  - Kayabaş2026
ER  -