Search papers, labs, and topics across Lattice.
This study introduces SGNet, a lightweight convolutional architecture tailored for hyperspectral imaging (HSI) to classify fish freshness by effectively separating spectral and spatial feature extraction. By employing a dual attention mechanism, SGNet enhances the model's ability to focus on relevant features, achieving an impressive 97.8% classification accuracy with only 4.75 million parameters. The results highlight the model's efficiency and effectiveness, demonstrating a significant reduction in parameters compared to traditional architectures while maintaining high performance on a new dataset of refrigerator-stored salmon fillets.
SGNet achieves near-perfect freshness classification with a fraction of the parameters of conventional models, paving the way for real-time applications in food safety.
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.
