About

FoodAI™ offers the cutting-edge food image recognition technologies for advancing AI in food and healthcare, particularly for Singapore multi-racial multi-cultural food. The FoodAI engine is a multi-agent system including a multi-task neural network for dish and ingredient recognition, a cross-modal network for recipe retrieval, some specialized networks for mixed dishes recognition (e.g., economic rice and bee hoon), and a large visual-language model. The engine can recognize dish names, ingredients, nutrition facts, and calories. The food image recognition technology is powered by the state-of-the-art deep learning and generative AI techniques. FoodAI™ is developed by the R&D team led by Prof. Ngo Chong-Wah at the School of Computing and Information Systems (SCIS), Singapore Management University (SMU), Singapore.

FoodAI™ covers over 1,000 dishes (including drink and fruit) and 800 ingredients.