Accurately Identify Food with Machine Vision
Advanced Food Image Recognition Technology By Deep Learning
Tags: National University of Singapore, Singapore, Agriculture & Food, Wellness, Safety & Fitness
The ATNet deep learning approach enhances food recognition in photos by using privileged information and knowledge graphs. It aligns image features with descriptive data for accurate classification, outperforming other algorithms in tests. This technology is versatile, supporting any CNN model, and is ideal for health, F&B, and lifestyle apps, offering precise nutrition tracking and market insights.
IP Type or Form Factor: Software & Algorithm
TRL: 4 - minimum viable product built in lab
Industry or Tech Area: Food & Beverage; Health & Wellness
