Meal Description
Meal description analysis is a burgeoning field focusing on automatically understanding and interpreting information about meals, encompassing image recognition, textual descriptions, and sensor data. Current research employs machine learning techniques, including clustering algorithms (like K-means and GMM) for sensor data analysis, deep learning models for image-based food identification, and large language models (LLMs) for nutritional estimation from textual descriptions. This work aims to improve dietary assessment, personalized nutrition recommendations, and diabetes management, ultimately impacting both public health initiatives and the development of more effective AI-driven healthcare tools.
Papers
Meal-taking activity monitoring in the elderly based on sensor data: Comparison of unsupervised classification methods
Abderrahim Derouiche (LAAS-S4M, UT3), Damien Brulin (LAAS-S4M, UT2J), Eric Campo (LAAS-S4M, UT2J), Antoine Piau
Detecting Korean Food Using Image using Hierarchical Model
Hoang Khanh Lam, Kahandakanaththage Maduni Pramuditha Perera