The Korean Journal of Food And Nutrition (한국식품영양학회지)

The Korean Society of Food and Nutrition (한국식품영양학회)
권호 표지
DOI QR코드

DOI QR Code

Development of an Optimal Convolutional Neural Network Backbone Model for Personalized Rice Consumption Monitoring in Institutional Food Service using Feature Extraction

  • Received : 2024.07.22
  • Accepted : 2024.08.14
  • Published : 2024.08.31
Download PDF
⟨ Previous Next ⟩

Abstract

This study aims to develop a deep learning model to monitor rice serving amounts in institutional foodservice, enhancing personalized nutrition management. The goal is to identify the best convolutional neural network (CNN) for detecting rice quantities on serving trays, addressing balanced dietary intake challenges. Both a vanilla CNN and 12 pre-trained CNNs were tested, using features extracted from images of varying rice quantities on white trays. Configurations included optimizers, image generation, dropout, feature extraction, and fine-tuning, with top-1 validation accuracy as the evaluation metric. The vanilla CNN achieved 60% top-1 validation accuracy, while pre-trained CNNs significantly improved performance, reaching up to 90% accuracy. MobileNetV2, suitable for mobile devices, achieved a minimum 76% accuracy. These results suggest the model can effectively monitor rice servings, with potential for improvement through ongoing data collection and training. This development represents a significant advancement in personalized nutrition management, with high validation accuracy indicating its potential utility in dietary management. Continuous improvement based on expanding datasets promises enhanced precision and reliability, contributing to better health outcomes.

Keywords

Acknowledgement

This work was supported by 2024 Bucheon University Research Grant.

References

  1. Bolanos M, Radeva P. 2016. Simultaneous food localization and recognition. In 2016 23rd International Conference on Pattern Recognition (ICPR). pp.3140-3145. IEEE
  2. Cai Q, Li J, Li H, Weng Y. 2019. BTBUFood-60: Dataset for object detection in food field. Available from https://doi.org/10.1109/BIGCOMP.2019.8678916 [cited 20 March 2024]
  3. Chen LC, Papandreou G, Kokkinos I, Murphy K, Yuille AL. 2017. Deeplab: Semantic image segmentation with deep convolution nets, atrous convolutional, and fully connected CRFs. IEEE Trans Pattern Anal Mach Intell 40:834-848
  4. Chollet F. 2017. Xception: Deep learning with depthwise separable convolution. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. pp.1251-1258. IEEE
  5. Fragopoulou E, Detopoulou P, Alepoudea E, Nomikos T, Kalogeropoulos N, Antonopoulou S. 2021. Associations between red blood cells fatty acids, desaturases indices and metabolism of platelet activating factor in healthy volunteers. Prostaglandins Leukot Essent Fatty Acids 164:102234
  6. Girshick R, Donahue J, Darrell T, Malik J. 2014. Rich feature hierarchies for accurate object detection and semantic segmentation. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. pp.580-587. IEEE
  7. He K, Zhang X, Ren S, Sun J. 2016. Deep residual learning for image recognition. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. pp.770-778. IEEE
  8. Hou W, Han T, Sun X, Chen Y, Xu J, Wang Y, Yang X, Jiang W, Sun C. 2022. Relationship between carbohydrate intake (quantity, quality, and time eaten) and mortality (total, cardiovascular, and diabetes): Assessment of 2003-2014 National Health and Nutrition Examination Survey participants. Diabetes Care 45:3024-3031
  9. Howard A, Sandler M, Chu G, Chen LC, Chen B, Tan M, Wang W, Zhu Y, Pang R, Vasudevan V, Le QV, Adam H. 2019. Searching for MobileNetV3. Available from https://arxiv.org/abs/1905.02244 [cited 20 March 2024]
  10. Howard A, Zhu M, Chen B, Kalenichenko D, Wang W, Weyand T, Andreetto M, Adam H. 2017. MobileNets: Efficient convolutional neural networks for mobile vision applications. arXiv preprint. Available from https://arxiv.org/abs/1704.04861 [cited 20 March 2024]
  11. Huang G, Liu Z, Van Der Maaten L, Weinberger KQ. 2017. Densely connected convolutional networks. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR). pp.4700-4708. IEEE
  12. Kalivaraprasad B, Prasad MVD, Vamsi R, Tejasri U, Santhoshi MN, PramodKumar A. 2021. Analysis of food recognition and calorie estimation using AI. In 1st International Conference on Advances in Signal Processing, Vlsi, Communications and Embedded Systems: Icsvce-2021 p.020020. AIP Publishing
  13. Kawano Y, Yanai K. 2014. Food image recognition with deep convolutional features. In UbiComp '14 Adjunct: Proceedings of the 2014 ACM International Joint Conference on Pervasive and Ubiquitous Computing. pp.589-593. Association for Computing Machinery
  14. Kim MH. 2013. Characteristics of nutrient intake according to metabolic syndrome in Korean elderly-using data from the Korea national health and nutrition examination survey 2010. Korean J Food Nutr 26:515-525
  15. Kingma DP, Ba J. 2014. Adam: A method for stochastic optimization. Available from https://arxiv.org/abs/1412.6980 [cited 20 March 2024]
  16. Korean Nutrition Society. 2020. Dietary reference intakes for Koreans 2020. Available from https://www.kns.or.kr/fileroom/fileroom.asp?BoardID=Kdr [cited 20 March 2024]
  17. Krizhevsky A, Sutskever I, Hinton GE. 2017. ImageNet classification with deep convolutional neural networks. Commun ACM 60:84-90
  18. Kurbiel T, Khaleghian S. 2017. Training of deep neural networks based on distance measures using RMSProp. Available from https://arxiv.org/abs/1708.01911 [cited 20 March 2024]
  19. Lin Y, Lv F, Zhu S, Yang M, Cour T, Yu K, Cao L, Huang T. 2011. Large-scale image classification: Fast feature extraction and SVM training. In 2011 IEEE Conference on on Computer Vision and Pattern Recognition CVPR 2011. pp.1689-1696. IEEE
  20. Liu W, Anguelov D, Erhan D, Szegedy C, Reed S, Fu CY, Berg AC. 2016. SSD: Single shot multibox detector. In Computer Vision - ECCV 2016: 14th European Conference on Computer Vision. pp.21-37. Springer
  21. Lu Y, Stathopoulou T, Vasiloglou MF, Pinault LF, Kiley C, Spanakis EK, Mougiakakou S. 2020. goFOODTM: An artificial intelligence system for dietary assessment. Sensors 20:4283
  22. Matusheski NV, Caffrey A, Christensen L, Mezgec S, Surendran S, Hjorth MF, McNulty H, Pentieva K, Roager HM, Seljak BK, Vimaleswaran KS, Remmers M, Peter S. 2021. Diets, nutrients, genes and the microbiome: Recent advances in personalised nutrition. Br J Nutr 126:1489-1497
  23. Mezgec S, Korousic Seljak B. 2017. NutriNet: A deep learning food and drink image recognition system for dietary assessment. Nutrients 9:657
  24. National Academies of Sciences, Engineering, and Medicine. 2016. Review of WIC Food Packages: Proposed Framework for Revisions: Interim Report. The National Academies Press
  25. Pan B, Zhao N, Xie Q, Li Y, Hamaker BR, Miao M. 2023. Molecular structure and characteristics of phytoglycogen, glycogen and amylopectin subjected to mild acid hydrolysis. npj Sci Food 7:27
  26. Peano C, Girgenti V, Sciascia S, Barone E, Sottile F. 2022. Dietary patterns at the individual level through a nutritional and environmental approach: The case study of a school canteen. Foods 11:1008
  27. Ronneberger O, Fischer P, Brox T. 2015. U-net: Convolutional networks for biomedical image segmentation. In Medical Image Computing and Computer-Assisted Intervention: MICCAI 2015. pp.234-241. Springer
  28. Sahoo D, Hao W, Ke S, Xiongwei W, Le H, Achananuparp P, Lim E, Hoi SC. 2019. FoodAI: Food image recognition via deep learning for smart food logging. In Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining. pp.2260-2268. Association for Computing Machinery
  29. Sandler M, Howard AG, Zhu M, Zhmoginov A, Chen LC. 2018. MobileNetV2: Inverted residuals and linear bottlenecks. Available from https://arxiv.org/abs/1801.04381 [cited 20 March 2024]
  30. Santamarina AB, Mennitti LV, de Souza EA, de Souza Mesquita LM, Noronha IH, Vasconcelos JRC, Prado CM, Pisani LP. 2023. A low-carbohydrate diet with different fatty acids' sources in the treatment of obesity: Impact on insulin resistance and adipogenesis. Clin Nutr 42:2381-2394
  31. Scherer D, Muller A, Behnke S. 2010. Evaluation of pooling operations in convolutional architectures for object recognition. In Artificial Neural Networks - ICANN 2010. pp.92-101. Springer
  32. Simonyan K, Zisserman A. 2015. Very deep convolutional networks for large-scale image recognition. Available from https://arxiv.org/abs/1409.1556 [cited 20 March 2024]
  33. Srivastava N, Hinton G, Krizhevsky A, Sutskever I, Salakhutdinov R. 2014. Dropout: a simple way to prevent neural networks from overfitting. J Mach Learn Res 15:1929-1958
  34. Stubbs RJ. 2021. Impact of carbohydrates, fat and energy density on energy intake. Nat Med 27:200-201
  35. Szegedy C, Liu W, Jia Y, Sermanet P, Reed S, Anguelov D, Erhan D, Vanhoucke V, Rabinovich A. 2015. Going deeper with convolution. In 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). pp.1-9. IEEE
  36. Tan M, Le QV. 2019. EfficientNet: Rethinking model scaling for convolutional neural networks. Available from https://arxiv.org/abs/1905.11946 [cited 20 March 2024]
  37. Vasiloglou MF, Mougiakakou S, Aubry E, Bokelmann A, Fricker R, Gomes F, Guntermann C, Meyer A, Studerus D, Stanga Z. 2018. A comparative study on carbohydrate estimation: GoCARB vs. dietitians. Nutrients 10:741
  38. Wali JA, Ni D, Facey HJW, Dodgson T, Pulpitel TJ, Senior AM, Raubenheimer D, Macia L, Simpson SJ. 2023. Determining the metabolic effects of dietary fat, sugars and fat-sugar interaction using nutritional geometry in a dietary challenge study with male mice. Nat Commun 14:4409
  39. Yang Q, Lang X, Li W, Liang Y. 2022. The effects of low-fat, high-carbohydrate diets vs. low-carbohydrate, high-fat diets on weight, blood pressure, serum liquids and blood glucose: A systematic review and meta-analysis. Eur J Clin Nutr76:16-27
  40. Yeom MY, Choi EY. 2023. Correlation of the nutrition quotient between parents and picky eaters in preschoolers. Korean J Food Nutr 36:103-113