Food Recognition Enhanced Using Privileged Information

25 May, 2020 by Jonathan Staniforth

Patent Name: Food Recognition Enhanced Using Privileged Information

Filing No: 10201907991T (SG)

Filing Date: 29 Aug 2019

Country to be Filed: SG

Description: This invention, called alignment and transfer network (ATNet), is a deep learning-based approach for food recognition from consumer photos. In addition to the conventional approaches based on solely image classification, it incorporates privileged information (PI) of food images for enhanced recognition power. PI refers to any descriptive information for the corresponding food images, such as ingredients, which is more discriminative than visual content. ATNet achieves this by enabling a two-stage cross-modal information passing. In the first stage, ATNet learns to align the feature representations of images to those of PI to take advantage of PI’s stronger discriminative power. In the second stage, images are mapped to recover PI for food classification. This makes the final decision a multi-view fusion of the classification results from the two stages

Patents

QLive: Low Latency Live Streaming Solution

25 May, 2020

Advanced Conversational Recommender System

25 May, 2020

Multi-Channel Graph Neural Network (MuGNN)

25 May, 2020

Food Recognition Enhanced Using Privileged Information

25 May, 2020

Knowledge Enhanced Translation-based User Preference Model (KTUP)

25 May, 2020

A Contextual Relation Networks for Mixed-Dish Recognition

25 May, 2020

Tree Enhanced Embedding Model Predictive Analysis Methods and Systems

25 May, 2020

Machine Learning Using Partial Order Hypergraphs

25 May, 2020

Video Visual Relation Detection

25 May, 2020

Object Trajectory Proposal

25 May, 2020

Neural Factorization Machines for Predictive Analytics

25 May, 2020

Attentional Factorization Machines: Learning the Weight of Feature Interactions via Attention Networks

22 May, 2020

Back