Hierarchical Spatio-Channel Clustering for Efficient Model Compression in Medical Image Analysis

TL;DR

This paper introduces a hierarchical spatio-channel low-rank compression framework for CNNs that exploits local structure in feature maps, significantly reducing computation while improving accuracy in medical image classification.

Contribution

The proposed method uniquely partitions feature maps into regions and groups channels based on co-activation, applying adaptive SVD to enhance compression efficiency and model performance.

Findings

Reduces FLOPs from 8.21G to 1.55G, an 81.1% reduction.

Achieves 1.38× inference speed-up and improves accuracy from 87.76% to 89.80%.

Outperforms global SVD and Tucker decomposition baselines in medical image classification.

Abstract

Convolutional neural networks (CNNs) have become increasingly difficult to deploy in resource-constrained environments due to their large memory and computational requirements. Although low-rank compression methods can reduce this burden, most existing approaches compress spatial and channel redundancy independently and therefore do not fully exploit the localised structure within convolutional feature maps. This paper proposes a hierarchical spatio-channel low-rank compression framework for CNNs that exploits redundancy across spatial regions and channel activations. Unlike conventional methods, which apply a uniform decomposition across an entire layer, the proposed approach first partitions feature maps into spatial regions, then groups channels according to their co-activation patterns within each region, and finally applies rank-adaptive SVD to each resulting spatio-channel cluster. The method is evaluated on an AlexNet-based brain tumour MRI classification model and compared with Global SVD and Tucker decomposition under \(3\times\) and \(6\times\) compression budgets. Our method outperforms both baselines, reducing FLOPs from \(8.21\,\mathrm{G}\) to \(1.55\,\mathrm{G}\) (\(81.1\%\) reduction), achieving a \(1.38\times\) inference speed-up, and increasing classification accuracy from \(87.76\%\) to \(89.80\%\). The method also improves the macro \(F_1\)-score and performance on challenging classes such as meningioma. A hyper-parameter trade-off analysis demonstrates that the framework provides Pareto-optimal configurations, enabling control over the balance between compression and predictive performance. Moderate clustering with adaptive rank selection yields strong results. Bootstrap standard errors are reported for all classification metrics.