AbstractAccurate and effortless nutritional tracking is a cornerstone of modern health and wellness, yet current methods are often tedious, inaccurate, and fail to handle complex, home-cooked meals. This paper presents a novel nutrition analysis system that leverages the power of Generative Artificial Intelligence (Gen AI) to overcome these limitations. The proposed system utilizes a state-of-the-art Large Multimodal Model (LMM) to perform a deep semantic analysis of food images. Unlike traditional systems that merely classify food items, our Gen AI-based approach generates a detailed, structured breakdown of a meal, including inferred ingredients, estimated portion sizes, and cooking methods. This generated data is then used to perform a precise nutritional calculation. Furthermore, the system employs a Large Language Model (LLM) to deliver personalized, conversational feedback and actionable recommendations to the user. This framework represents a paradigm shift from simple data logging to an interactive, intelligent nutritional partner, making personalized health management more accessible and effective. Keywords—Generative AI, Nutrition Analysis, Large Multimodal Models (LMM), Computer Vision, Personalized Health, Natural Language Generation (NLG), Food Recognition
M et al. (Tue,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: