A groundbreaking AI system developed by researchers from NYU Tandon School of Engineering is poised to revolutionize how we track our diet. With a simple snap of a meal, the system instantly calculates the nutritional values—calories, protein, fat, carbohydrates—without the need for food diaries or manual input. This technology could be a game-changer for people managing weight, diabetes, or other diet-related health issues.
The system, which uses advanced deep-learning algorithms, was designed to tackle long-standing challenges in food recognition and dietary tracking. One major hurdle was the vast diversity of food—each dish can look dramatically different depending on who prepares it. Previous attempts at food-recognition AI struggled with accurately identifying and quantifying different types of food and their portions. The NYU team’s breakthrough lies in their volumetric computation function, which uses image processing to measure the area occupied by food on a plate, offering precise nutritional calculations.
The system runs in real-time and can be accessed through a website, bypassing the need for apps that may require more processing power or introduce privacy concerns. For instance, when tested on a pizza slice, the system calculated 317 calories, 10 grams of protein, 40 grams of carbohydrates, and 13 grams of fat—results that closely matched reference standards.
Currently available as a proof-of-concept, this system is being refined for broader health care applications and could soon be a daily tool for anyone looking to track their meals and nutrition effortlessly.
Read also;