Deep feature optimization for enhanced fish freshness assessment

Assessing fish eye freshness is vital for ensuring food safety and minimizing economic losses in the seafood industry. However, traditional sensory evaluation methods remain subjective, time-consuming, and inconsistent. Despite recent advancements in deep learning for automating visual freshness prediction, challenges related to accuracy and feature transparency persist. This study introduces a unified three-stage framework that refines and leverages deep visual representations for reliable fish eye freshness assessment. First, five state-of-the-art vision architectures–ResNet-50, DenseNet121, EfficientNet-B0, ConvNeXt-Base, and Swin-Tiny–are fine-tuned to establish a strong baseline. Next, multi-level deep features extracted from these backbones are used to train seven classical machine learning classifiers, integrating deep and traditional decision mechanisms. Finally, feature selection methods based on Light Gradient Boosting Machine (LGBM), Random Forest, and Lasso are utilized to identify a compact and informative subset of features. Experiments conducted on the Freshness of the Fish Eyes (FFE) dataset demonstrate that the best configuration–combining Swin-Tiny features, a Random Forest classifier, and LGBM-based feature selection–achieves an accuracy of 85.99%, outperforming recent studies on the same dataset by 8.69–22.78%. These findings confirm the effectiveness and generalizability of the proposed framework for visual quality evaluation tasks. • A unified three-stage framework is proposed for automated fish eye freshness assessment. • Multi-level deep feature extraction and embedded feature selection enhance discriminative power. • Deep features from five state-of-the-art architectures are optimized via hybrid learning. • The Swin-Tiny + Random Forest + LGBM-based configuration achieves an accuracy of 85.99%. • The proposed framework outperforms existing methods on the same dataset by 8.69%–22.78%.