Technology
ChironSoft does its best from the customer's perspective.
A learning model learned with the TensorFlow deep learning framework. It receives the user's food image as input, performs analysis, learning, and model creation, and provides food recognition results to the user.
System Configuration
The food recognition application implements personalized food recognition technology and applies deep learning technology based on CNN (Convolutional Neural Network) to learn, infer, and classify food images. Personalized food recognition technology increases the rate of food recognition by assigning weight to foods that an individual enjoys eating.
Food image learning and inference process
The deep learning framework for food recognition used Tensorflow and a network structure called Google InceptionV3. In this project, food image learning and model creation are performed on the server, and the mobile application receives the learning model and performs food recognition (classification) work.
(Scalars)-TensorBoard
(Histograms)-TensorBoard
