Improved food recognition using a refined ResNet50 architecture with improved fully connected layers
Food consumption has significant effects on human health, particularly in relation to quality control, quantity, freshness, and color. This study focuses on identifying food across 16 categories, divided into breakfast, lunch, and dinner, to evaluate its impact on the human body, specifically in hos...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2025-01-01
|
| Series: | Current Research in Food Science |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S266592712500036X |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850102918503464960 |
|---|---|
| author | Pouya Bohlol Soleiman Hosseinpour Mahmoud Soltani Firouz |
| author_facet | Pouya Bohlol Soleiman Hosseinpour Mahmoud Soltani Firouz |
| author_sort | Pouya Bohlol |
| collection | DOAJ |
| description | Food consumption has significant effects on human health, particularly in relation to quality control, quantity, freshness, and color. This study focuses on identifying food across 16 categories, divided into breakfast, lunch, and dinner, to evaluate its impact on the human body, specifically in hospital and restaurant settings. The recognition system was used a machine vision system and deep learning algorithms to record food consumption videos, extracting images. After preprocessing, a raw dataset was bulit that consisted of 12,000 images, expanded to 66,000 images through data augmentation. Five deep learning algorithms were used for recognizing food and consumed food. ResNet50 was the best algorithm in comparison to other deep learning architectures. The effect of Hyper-parameters such as data augmentation, batch size, image size, and learning rate on performance of Resnet50 were analyzed. Transfer learning method led us to develop three versions: standard ResNet50, fine-tuned ResNet50, and optimized ResNet50 with a customized fully connected layer. ResNet50 with a specific dense layer was the best development version of ResNet50. This model with Adam optimizer, 10−3 initial learning rate, batch size 4, and image size 340 × 640 could recognize various foods with 97.25% accuracy and 0.2 loss. Response time and training time of this architecture compared to other algorithms were confidential; the training process and response time were 5.30 h and 1.2 s. ResNet50 with a specific fully connected layer powerfully could complete tasks with high accuracy and the least time. |
| format | Article |
| id | doaj-art-40c2111d08c44db890558d126e1a020a |
| institution | DOAJ |
| issn | 2665-9271 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | Elsevier |
| record_format | Article |
| series | Current Research in Food Science |
| spelling | doaj-art-40c2111d08c44db890558d126e1a020a2025-08-20T02:39:39ZengElsevierCurrent Research in Food Science2665-92712025-01-011010100510.1016/j.crfs.2025.101005Improved food recognition using a refined ResNet50 architecture with improved fully connected layersPouya Bohlol0Soleiman Hosseinpour1Mahmoud Soltani Firouz2Department of Agricultural Machinery Engineering, Faculty of Agricultural Engineering and Technology, University of Tehran, Karaj, IranCorresponding author.; Department of Agricultural Machinery Engineering, Faculty of Agricultural Engineering and Technology, University of Tehran, Karaj, IranDepartment of Agricultural Machinery Engineering, Faculty of Agricultural Engineering and Technology, University of Tehran, Karaj, IranFood consumption has significant effects on human health, particularly in relation to quality control, quantity, freshness, and color. This study focuses on identifying food across 16 categories, divided into breakfast, lunch, and dinner, to evaluate its impact on the human body, specifically in hospital and restaurant settings. The recognition system was used a machine vision system and deep learning algorithms to record food consumption videos, extracting images. After preprocessing, a raw dataset was bulit that consisted of 12,000 images, expanded to 66,000 images through data augmentation. Five deep learning algorithms were used for recognizing food and consumed food. ResNet50 was the best algorithm in comparison to other deep learning architectures. The effect of Hyper-parameters such as data augmentation, batch size, image size, and learning rate on performance of Resnet50 were analyzed. Transfer learning method led us to develop three versions: standard ResNet50, fine-tuned ResNet50, and optimized ResNet50 with a customized fully connected layer. ResNet50 with a specific dense layer was the best development version of ResNet50. This model with Adam optimizer, 10−3 initial learning rate, batch size 4, and image size 340 × 640 could recognize various foods with 97.25% accuracy and 0.2 loss. Response time and training time of this architecture compared to other algorithms were confidential; the training process and response time were 5.30 h and 1.2 s. ResNet50 with a specific fully connected layer powerfully could complete tasks with high accuracy and the least time.http://www.sciencedirect.com/science/article/pii/S266592712500036XRefined ResNet50Specific connected layersOptimized hyperparamtersIranian foodDeep learning |
| spellingShingle | Pouya Bohlol Soleiman Hosseinpour Mahmoud Soltani Firouz Improved food recognition using a refined ResNet50 architecture with improved fully connected layers Current Research in Food Science Refined ResNet50 Specific connected layers Optimized hyperparamters Iranian food Deep learning |
| title | Improved food recognition using a refined ResNet50 architecture with improved fully connected layers |
| title_full | Improved food recognition using a refined ResNet50 architecture with improved fully connected layers |
| title_fullStr | Improved food recognition using a refined ResNet50 architecture with improved fully connected layers |
| title_full_unstemmed | Improved food recognition using a refined ResNet50 architecture with improved fully connected layers |
| title_short | Improved food recognition using a refined ResNet50 architecture with improved fully connected layers |
| title_sort | improved food recognition using a refined resnet50 architecture with improved fully connected layers |
| topic | Refined ResNet50 Specific connected layers Optimized hyperparamters Iranian food Deep learning |
| url | http://www.sciencedirect.com/science/article/pii/S266592712500036X |
| work_keys_str_mv | AT pouyabohlol improvedfoodrecognitionusingarefinedresnet50architecturewithimprovedfullyconnectedlayers AT soleimanhosseinpour improvedfoodrecognitionusingarefinedresnet50architecturewithimprovedfullyconnectedlayers AT mahmoudsoltanifirouz improvedfoodrecognitionusingarefinedresnet50architecturewithimprovedfullyconnectedlayers |