Applied Intelligence and Computing
Industry 4.0 Meets Gastronomy: Elevating Pandang Cuisine Sorting with Cutting-Edge Transfer Learning
Authors: Mohit Beri, Kanwarpartap Singh Gill, Deepak Upadhyay and Swati Devliyal
Publishing Date: 05-11-2024
ISBN: 978-81-955020-9-7
Abstract
Deep Convolutional Neural Networks (CNNs) and Transfer Learning are examined in this study article as they pertain to the classification of Pandang cuisine. The goal is to use AI technology to make the food and drink industry more efficient and profitable. Recognising the significance of Industry 4.0, the study employed a dataset including 993 images of 9 different Pandang dishes. The technique involves loading, converting, and preparing the data using the MobileNetV2 pre-trained CNN model. Then, the model is trained utilising callbacks such as Model Checkpoint and Early Stopping. The hyperparameters include a batch size of 32, an output layer with 9 classes, and 100 epochs. Using other evaluation metrics such as F1 score, recall, and precision, the model achieves a remarkable accuracy of 90% on the test dataset. Displayed with Grad-Cam visualisations, the study concludes with accuracy and loss curves, test data projections, Classification Reports, and a Confusion Matrix. This research has the makings of a game-changing tool for the food industry's supply chain management and culinary procedures.
Keywords
Artificial Intelligence, Deep Learning, Pandang Cuisine Classification, Model Training, MobileNetV2 CNN Model.
Cite as
Mohit Beri, Kanwarpartap Singh Gill, Deepak Upadhyay and Swati Devliyal, "Industry 4.0 Meets Gastronomy: Elevating Pandang Cuisine Sorting with Cutting-Edge Transfer Learning", In: Mukesh Saraswat and Rajani Kumari (eds), Applied Intelligence and Computing, SCRS, India, 2024, pp. 57-64. https://doi.org/10.56155/978-81-955020-9-7-7