DASS-Net: A Multi-Branch Aesthetic–Spectral–Semantic Deep Learning Model for Food Spoilage Detection in Hospitality Operations
PDFAbstract
Food freshness and food safety in the hospitality industry are, to date, mainly dependent on visual observation, where human visual interpretation does not always successfully identify the less obvious signs of early food spoilage. In this work, we introduce DASS-Net, a novel hybrid deep network with a trilinear structure and three separate task-focused branches: (1) the RGB aesthetic branch, (2) the LAB spectral branch, and (3) the Vision Transformer ViT-Small semantic branch, used together for simultaneous analysis of food image-based texture degradation, color deviation, and overall food image-related semantic dependencies on a global scale. Each of the network's task-focused branches projects its outputs to a unitary latent space and further refines them with multi-head attention and adaptive joint processing with a learnable gated fusion module. To train DASS-Net, a dual-objective training approach with class-weighted and smoothed Cross-Entropy loss and Contrastive Learning with MixUp and CutMix data augmentations is used. Eight-class classification experiments on fresh and spoiled food samples of bread, dairy products, fruits, and vegetables are conducted to evaluate the performance of DASS-Net, providing a validation accuracy of 94.33%, which demonstrates a 23.18% and 25.83% relative improvement over ResNet-18 with LAB and RGB visual channels, respectively. Additionally, the classification model yielded a minimum misclassification rate of 13.30% for spoiled items identified as fresh, which clearly marks an important characteristic related to safety issues. The complementarity between the spectral and aesthetic features has also been proven by the Grad-CAM visualizations of the dual-branch network, where the most confusion occurs in the case of dairy products with negligible surface decay, as revealed by the analysis of failure cases. The results explicitly confirm that the developed DASS-Net delivers a safe, efficient, and scalable vision solution applicable to kitchen intelligence and hospitality control scenarios.
Cite this article
Descarten, Randy O., and Jasten Keneth D. Treceñe. "DASS-Net: A Multi-Branch Aesthetic–Spectral–Semantic Deep Learning Model for Food Spoilage Detection in Hospitality Operations." Journal of Innovative Image Processing 7, no. 4 (2025): 1482-1506. doi: 10.36548/jiip.2025.4.021
Copy Citation
Descarten, R. O., & Treceñe, J. K. D. (2025). DASS-Net: A Multi-Branch Aesthetic–Spectral–Semantic Deep Learning Model for Food Spoilage Detection in Hospitality Operations. Journal of Innovative Image Processing, 7(4), 1482-1506. https://doi.org/10.36548/jiip.2025.4.021
Copy Citation
Descarten, Randy O., et al. "DASS-Net: A Multi-Branch Aesthetic–Spectral–Semantic Deep Learning Model for Food Spoilage Detection in Hospitality Operations." Journal of Innovative Image Processing, vol. 7, no. 4, 2025, pp. 1482-1506. DOI: 10.36548/jiip.2025.4.021.
Copy Citation
Descarten RO, Treceñe JKD. DASS-Net: A Multi-Branch Aesthetic–Spectral–Semantic Deep Learning Model for Food Spoilage Detection in Hospitality Operations. Journal of Innovative Image Processing. 2025;7(4):1482-1506. doi: 10.36548/jiip.2025.4.021
Copy Citation
R. O. Descarten, and J. K. D. Treceñe, "DASS-Net: A Multi-Branch Aesthetic–Spectral–Semantic Deep Learning Model for Food Spoilage Detection in Hospitality Operations," Journal of Innovative Image Processing, vol. 7, no. 4, pp. 1482-1506, Dec. 2025, doi: 10.36548/jiip.2025.4.021.
Copy Citation
Descarten, R.O. and Treceñe, J.K.D. (2025) 'DASS-Net: A Multi-Branch Aesthetic–Spectral–Semantic Deep Learning Model for Food Spoilage Detection in Hospitality Operations', Journal of Innovative Image Processing, vol. 7, no. 4, pp. 1482-1506. Available at: https://doi.org/10.36548/jiip.2025.4.021.
Copy Citation
@article{descarten2025,
author = {Randy O. Descarten and Jasten Keneth D. Treceñe},
title = {{DASS-Net: A Multi-Branch Aesthetic–Spectral–Semantic Deep Learning Model for Food Spoilage Detection in Hospitality Operations}},
journal = {Journal of Innovative Image Processing},
volume = {7},
number = {4},
pages = {1482-1506},
year = {2025},
publisher = {Inventive Research Organization},
doi = {10.36548/jiip.2025.4.021},
url = {https://doi.org/10.36548/jiip.2025.4.021}
}
Copy Citation
Keywords
Food Spoilage Detection
Hospitality Management
Vision Transformer
Spectral–Aesthetic Fusion
Supervised Contrastive Learning
References
- Elimelech, Efrat, Eyal Ert, Yael Parag, and Guy Hochman. "Exploring The Impact of Visual Perception and Taste Experience on Consumers’ Acceptance of Suboptimal Fresh Produce." Sustainability 16, no. 7 (2024): 2698.
- Okpala, Charles Odilichukwu R., and Małgorzata Korzeniowska. "Understanding the Relevance of Quality Management in Agro-Food Product Industry: From Ethical Considerations to Assuring Food Hygiene Quality Safety Standards and Its Associated Processes." Food Reviews International 39, no. 4 (2023): 1879-1952.
- Creed, P. G. "Sensory Quality Control in Foodservice." Sensory Analysis for Food and Beverage Quality Control (2010): 316-336.
- Cohen Hakmon, May, Keren Buhnik-Rosenblau, Hila Hanani, Hila Korach-Rechtman, Dagan Mor, Erez Etkin, and Yechezkel Kashi. "Early Detection of Food Safety and Spoilage Incidents Based on Live Microbiome Profiling and PMA-qPCR Monitoring of Indicators." Foods 13, no. 15 (2024): 2459.
- Segbedzi, Cynthia Esinam, Edward Wilson Ansah, and Daniel Apaak. "Compliance to Food Safety Standards: Determining the Barriers Within the Hotel Industry." PLOS Global Public Health 5, no. 11 (2025): e0002771.
- Balakrishnan, P., A. Anny Leema, N. Jothiaruna, Purshottam J. Assudani, K. Sankar, Madhusudan B. Kulkarni, and Manish Bhaiyya. "Artificial Intelligence for Food Safety: From Predictive Models to Real-World Safeguards." Trends in Food Science & Technology 163 (2025): 105153.
- Ahmad, Muhammad, Salvatore Distefano, Adil Mehmood Khan, Manuel Mazzara, Chenyu Li, Hao Li, Jagannath Aryal, Yao Ding, Gemine Vivone, and Danfeng Hong. "A Comprehensive Survey for Hyperspectral Image Classification: The Evolution from Conventional to Transformers and Mamba Models." Neurocomputing (2025): 130428.
- Anwar, Abbas, Saira Kanwal, Muhammad Tahir, Muhammad Saqib, Muhammad Uzair, Mohammad Khalid Imam Rahmani, and Habib Ullah. "Image Aesthetic Assessment: A Comparative Study of Hand-Crafted & Deep Learning Models." IEEE Access 10 (2022): 101770-101789.
- Pu, Yumei, Danfei Liu, Siyuan Chen, and Yunfei Zhong. "Research Progress on the Aesthetic Quality Assessment of Complex Layout Images Based on Deep Learning." Applied Sciences 13, no. 17 (2023): 9763.
- Treepong, Panisa, and Nawanol Theera-Ampornpunt. "Early Bread Mold Detection Through Microscopic Images Using Convolutional Neural Network." Current Research in Food Science 7 (2023): 100574.
- Jubayer, Fahad, Janibul Alam Soeb, Abu Naser Mojumder, Mitun Kanti Paul, Pranta Barua, Shahidullah Kayshar, Syeda Sabrina Akter, Mizanur Rahman, and Amirul Islam. "Detection of Mold on the Food Surface Using YOLOv5." Current Research in Food Science 4 (2021): 724-728.
- Feng, Chao-Hui. "Colour Analysis of Sausages Stuffed with Modified Casings Added with Citrus Peel Extracts Using Hyperspectral Imaging Combined with Multivariate Analysis." Sustainability 16, no. 19 (2024): 8683.
- Cheng, Lijuan, Guishan Liu, Jianguo He, Guoling Wan, Chao Ma, Jingjing Ban, and Limin Ma. "Non-Destructive Assessment of the Myoglobin Content of Tan Sheep Using Hyperspectral Imaging." Meat Science 167 (2020): 107988.
- Foong, Chai C., Goh K. Meng, and Lim L. Tze. "Convolutional Neural Network Based Rotten Fruit Detection Using Resnet50." In 2021 IEEE 12th Control and System Graduate Research Colloquium (ICSGRC), IEEE, 2021, 75-80.
- Gillespie, James, Jordan Vincent, Omar Dib, Matthias Heiden, and Joan Condell. "Handheld Spectroscopy for Dairy Quality Control: Towards a Real Time Milk Spoilage Classification Model." In International Conference on Ubiquitous Computing and Ambient Intelligence, Cham: Springer Nature Switzerland, 2024, 829-840.
- Khosla, Prannay, Piotr Teterwak, Chen Wang, Aaron Sarna, Yonglong Tian, Phillip Isola, Aaron Maschinot, Ce Liu, and Dilip Krishnan. "Supervised Contrastive Learning." Advances in neural information processing systems 33 (2020): 18661-18673.
- Fresh and Spoiled Food Image Dataset. (n.d.). Retrieved November 25, 2025, from https://www.kaggle.com/datasets/maheen00shahid/fresh-and-spoiled-food-image-dataset
- Govindarajan, Vijay, and Junaid Hussain Muzamal. "Advanced Cloud Intrusion Detection Framework Using Graph Based Features Transformers and Contrastive Learning." Scientific Reports 15, no. 1 (2025): 20511.
- Chen, Dongyu, and Haitao Zhao. "CCD-Net: Color-Correction Network Based on Dual-Branch Fusion of Different Color Spaces for Image Dehazing." Applied Sciences 15, no. 6 (2025): 3191.
- Hatamizadeh, Ali, Hongxu Yin, Greg Heinrich, Jan Kautz, and Pavlo Molchanov. "Global Context Vision Transformers." In International Conference on Machine Learning, pp. 12633-12646. PMLR, 2023.
- Salehi, Fakhreddin. "Quality, Physicochemical, And Textural Properties of Dairy Products Containing Fruits and Vegetables: A Review." Food Science & Nutrition 9, no. 8 (2021): 4666-4686.
- Lun, Zhichen, Xiaohong Wu, Jiajun Dong, and Bin Wu. "Deep Learning-Enhanced Spectroscopic Technologies for Food Quality Assessment: Convergence and Emerging Frontiers." Foods 14, no. 13 (2025): 2350.
- Nur Farzanah Faghira Kamarudin, Puteri, Nik Mohd Zarifie Hashim, Masrullizam Mat Ibrahim, and Mahmud Dwi Sulistiyo. "Milk Spoilage Classification Through Integration of RGB and Thermal Data Analysis." International Journal of Computing and Digital Systems 15, no. 1 (2024): 1839-1851.
- Kilcawley, Kieran N., Hope Faulkner, Holly J. Clarke, Maurice G. O’Sullivan, and Joseph P. Kerry. "Factors Influencing the Flavour of Bovine Milk and Cheese from Grass Based Versus Non-Grass Based Milk Production Systems." Foods 7, no. 3 (2018): 37.
- Practices of Science: False Positives and False Negatives | manoa.hawaii.edu/ExploringOurFluidEarth. (n.d.). Retrieved December 9, 2025, from https://manoa.hawaii.edu/exploringourfluidearth/chemical/matter/properties-matter/practices-science-false-positives-and-false-negatives
- Han, Xiaofeng, Shunpeng Chen, Zenghuang Fu, Zhe Feng, Lue Fan, Dong An, Changwei Wang et al. "Multimodal Fusion and Vision-Language Models: A survey for robot vision." Information Fusion (2025): 103652.
- Maurício, José, Inês Domingues, and Jorge Bernardino. "Comparing Vision Transformers and Convolutional Neural Networks for Image Classification: A Literature Review." Applied Sciences 13, no. 9 (2023): 5521.
- Yang, Yican, Nuwan K. Wijewardane, Lorin Harvey, and Xin Zhang. "Beyond Flips and Rotations: Evaluating NeRF and 3DGS-Based Synthetic View Augmentation: A Case Study of Detecting Visual Skinning Damage on Sweetpotato Storage Roots." Journal of Agriculture and Food Research (2025): 102517.
- Elkins, Andrew, Felipe F. Freitas, and Verónica Sanz. "Developing an App to Interpret Chest X-Rays to Support the Diagnosis of Respiratory Pathology with Artificial Intelligence." arXiv preprint arXiv:1906.11282 (2019).
- Araujo, Sufola Das Chagas Silva, Goh Kah Ong Michael, Uttam U. Deshpande, Sudhindra Deshpande, Manjunath G. Avalappa, Yash Amasi, Sumit Patil, Swathi Bhat, and Sudarshan Karigoudar. "ResNet-18 based Multi-Task Visual Inference and Adaptive Control for an Edge-Deployed Autonomous Robot." Frontiers in Robotics and AI 12 (2025): 1680285.
