轻量级网络精准识别鱼鲜度，仅用475万参数实现97.8%准确率。
Domain-Aware Lightweight Spectral-Grouped Convolutions for Hyperspectral Fish Freshness Classification
- 分谱与空间特征提取，用分组卷积和深度可分离路径降低计算负担。
- 在16天冷藏鲑鱼数据集上达97.8%准确率，均方误差仅0.64天。
- 适合工业部署，参数比ResNet-50少5到18倍，适合边缘设备使用。
高光谱成像（HSI）可通过光谱波段的生化变化实现鱼类新鲜度的无损评估。然而，传统深度学习方法未能充分考虑HSI数据的特性，如光谱主导性、有序标签结构及少量训练样本。本文提出SGNet（光谱分组网络），一种轻量级架构，通过分组卷积分离光谱与空间特征提取，并引入深度可分离空间路径。双重注意力机制结合通道挤压-激励与空间门控，自适应增强关键特征。在新构建的16天冷藏三文鱼片数据集上，SGNet仅用475万参数即达到97.8%分类准确率和0.64天均绝对误差。消融实验验证各组件贡献，对比表明其参数量较ResNet-50和视觉变换器减少五至十八倍。结果表明，领域感知设计有助于实现高精度、实时的工业级新鲜度预测。
原文摘要 · Abstract (English)
Hyperspectral imaging (HSI) offers nondestructive assessment of fish freshness by detecting biochemical alterations across spectral bands. However, conventional deep learning approaches do not fully address the particular characteristics of HSI data, such as spectral dominance over spatial textures, ordinal label structure, and a small number of training samples. We propose SGNet (Spectral-Grouped Network), a lightweight architecture that separates spectral and spatial feature extraction using grouped convolutions and a depthwise spatial pathway. A dual attention mechanism that couples channel-wise squeeze-and-excitation with spatial gating adaptively highlights informative features. SGNet achieves 97.8% classification accuracy and 0.64 days mean absolute error (MAE) with just 4.75M parameters when tested on our newly developed 16-day refrigerator-stored salmon fillet dataset. Ablation studies validate the contribution of each component, while comparisons demonstrate a five- to eighteen-fold parameter reduction relative to ResNet-50 and Vision Transformers. Our findings indicate that domain-aware design supports precise, real-time freshness prediction for industrial implementation.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。