Automated Meal Analysis and Realtime Dietary Recommender System: Comparative Analysis of Machine Learning Models for Benchmarking Performance and Innovation While Utilizing State-of-the-Art Algorithms for Caloric Content Recognition of Food for Alimentary Guidance in Support of Metabolic Disease Prevention and Management and of Promoting Overall Mindfulness of Fitness and Wellbeing
Guardado en:
| Publicado en: | ProQuest Dissertations and Theses (2025) |
|---|---|
| Autor principal: | |
| Publicado: |
ProQuest Dissertations & Theses
|
| Materias: | |
| Acceso en línea: | Citation/Abstract Full Text - PDF |
| Etiquetas: |
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| Resumen: | The global rise in noncommunicable diseases and aging populations accentuates the urgent need for proactive health management. One of the most important methods to achieve this is through dietary interventions, However, conventional dietary monitoring methods are mostly reliant on manual tracking, subjective recall, delayed data entry, and monitoring by human experts. Such methods can potentially struggle with inaccuracies, poor scalability, and low adherence. This research proposes an Automated Meal Analysis and Real-Time Dietary Recommender System. It employs state-of-the-art machine learning models to address these gaps. It compares the performance of two computer vision models, CLIP ViT-B/32, a vision transformer, and ResNet-50, a convolutional neural network, on a publicly available repository called Food-101, which is a curated, diverse dataset. The study evaluates the accuracy of the models in meal identification through food images. Both models achieved over 90% accuracy, with ResNet-50 demonstrating marginally superior performance. The performance of these models aligns with the high performance reported in other research utilizing cutting-edge deep learning models for food image recognition. This validates the hypothesis with which this research was embarked upon. It also stresses the promise of these models for integration into a real-time interactive platform, which would enable users to capture meal images for instant analysis and caloric estimation. The platform would provide personalized nutritional feedback to support health goals, chronic disease management, and preventive care. The applications of this research span clinical nutrition, fitness coaching, and food services, while the underlying vision techniques hold broader utility in fields requiring robust object recognition such as autonomous vehicles, crop disease identification and robotic vision. This research enhances automated meal assessment and bridges AI innovation with public health needs. |
|---|---|
| ISBN: | 9798293864904 |
| Fuente: | ProQuest Dissertations & Theses Global |