Nasi lemak calorie counter with deep neural network

This report presents the development of a Nasi Lemak Calorie Counter system using a deep learning approach. The objective of the project is to accurately detect and estimate the calorie content of various components in Nasi Lemak, a popular Malaysian dish. The methodology involves data collection, model training using a single-shot multibox detector (SSD) architecture, and integrating the trained model into a user-friendly interface. The system achieves accurate object detection and estimates calorie content based on the detected components. The performance of the system is evaluated using precision, recall, and mean Average Precision (mAP) metrics. The results show promising performance, with an overall mAP score of 32.27% across different components. The system's limitations are identified, including the need for a larger dataset and further optimization for real-time usage. Future directions include dataset expansion, integration of additional dishes, and enhancing real-time performance. The Nasi Lemak Calorie Counter system provides a valuable tool for individuals to monitor their calorie intake accurately and make informed dietary decisions.