<![CDATA[This study aims to develop an image-based recognition system for Indonesian traditional foods to support the preservation of Nusantara culinary heritage in the digital era using the Convolutional Neural Network (CNN) method. This research was conducted due to the very limited number of studies specifically focused on classifying Indonesian regional traditional foods. Using a transfer learning approach, the pre-trained ResNet50 model was employed as the main architecture, with fine-tuning applied to the final layers to adapt the classification to 13 categories of Indonesian traditional foods. The dataset consisted of 1,300 images that underwent preprocessing and data augmentation to enhance the model’s generalization performance. Evaluation results show that the model achieved an accuracy of 81%, with precision, recall, and F1-score values indicating strong classification performance across most classes. The model was integrated into a user interface system to support real-time image prediction. System testing demonstrated fast response capabilities and high prediction confidence. Overall, this study confirms that CNNs with transfer learning can serve as an effective solution for recognizing Indonesian traditional foods and hold potential for further development as an educational medium and a tool for promoting local culinary culture.]]>
Copyrights © 2025
