MAAM: A Lightweight Multi-Agent Aggregation Module for Efficient Image Classification Based on the MindSpore Framework

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Qin, Zhenkai, Zhu, Feng, Zeng, Huan, Nong, Xunyi
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910914039513088
author Qin, Zhenkai
Zhu, Feng
Zeng, Huan
Nong, Xunyi
author_facet Qin, Zhenkai
Zhu, Feng
Zeng, Huan
Nong, Xunyi
contents The demand for lightweight models in image classification tasks under resource-constrained environments necessitates a balance between computational efficiency and robust feature representation. Traditional attention mechanisms, despite their strong feature modeling capability, often struggle with high computational complexity and structural rigidity, limiting their applicability in scenarios with limited computational resources (e.g., edge devices or real-time systems). To address this, we propose the Multi-Agent Aggregation Module (MAAM), a lightweight attention architecture integrated with the MindSpore framework. MAAM employs three parallel agent branches with independently parameterized operations to extract heterogeneous features, adaptively fused via learnable scalar weights, and refined through a convolutional compression layer. Leveraging MindSpore's dynamic computational graph and operator fusion, MAAM achieves 87.0% accuracy on the CIFAR-10 dataset, significantly outperforming conventional CNN (58.3%) and MLP (49.6%) models, while improving training efficiency by 30%. Ablation studies confirm the critical role of agent attention (accuracy drops to 32.0% if removed) and compression modules (25.5% if omitted), validating their necessity for maintaining discriminative feature learning. The framework's hardware acceleration capabilities and minimal memory footprint further demonstrate its practicality, offering a deployable solution for image classification in resource-constrained scenarios without compromising accuracy.
format Preprint
id arxiv_https___arxiv_org_abs_2504_13574
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle MAAM: A Lightweight Multi-Agent Aggregation Module for Efficient Image Classification Based on the MindSpore Framework
Qin, Zhenkai
Zhu, Feng
Zeng, Huan
Nong, Xunyi
Machine Learning
Computer Vision and Pattern Recognition
Image and Video Processing
The demand for lightweight models in image classification tasks under resource-constrained environments necessitates a balance between computational efficiency and robust feature representation. Traditional attention mechanisms, despite their strong feature modeling capability, often struggle with high computational complexity and structural rigidity, limiting their applicability in scenarios with limited computational resources (e.g., edge devices or real-time systems). To address this, we propose the Multi-Agent Aggregation Module (MAAM), a lightweight attention architecture integrated with the MindSpore framework. MAAM employs three parallel agent branches with independently parameterized operations to extract heterogeneous features, adaptively fused via learnable scalar weights, and refined through a convolutional compression layer. Leveraging MindSpore's dynamic computational graph and operator fusion, MAAM achieves 87.0% accuracy on the CIFAR-10 dataset, significantly outperforming conventional CNN (58.3%) and MLP (49.6%) models, while improving training efficiency by 30%. Ablation studies confirm the critical role of agent attention (accuracy drops to 32.0% if removed) and compression modules (25.5% if omitted), validating their necessity for maintaining discriminative feature learning. The framework's hardware acceleration capabilities and minimal memory footprint further demonstrate its practicality, offering a deployable solution for image classification in resource-constrained scenarios without compromising accuracy.
title MAAM: A Lightweight Multi-Agent Aggregation Module for Efficient Image Classification Based on the MindSpore Framework
topic Machine Learning
Computer Vision and Pattern Recognition
Image and Video Processing
url https://arxiv.org/abs/2504.13574