HHSN261201400054C; TOPIC 308 AUTOMATED COLLECTION, STORAGE, ANALYSIS, AND REPORTING SYSTEMS FOR DIETARY IMAGES 'TITLE: THE MOBILE FOOD INTAKE PHOTO STORAGE & ANALYSIS SYSTEM. PERFORMANCE PERIOD 09/16/

点击翻译按钮获取中文摘要

This proposal describes enhancements to Viocare’s Mobile Food Intake Visual and Voice Recognizer (FIVR) System, a novel combination of innovative technologies including computer vision and speech recognition to measure dietary intake using a mobile phone. FIVR uses a mobile phone’s camera to capture a short video of foods to be consumed, which is then verbally-annotated on the mobile phone by the user. These video and audio files are processed through a real-time backend server speech and image recognition engine for food recognition and portion size measurement. This project will extend FIVR’s capabilities to analyze more foods, enhance the analysis and reporting tools, expand system support tools, and develop interfaces to a diverse set of clinical and research systems. A final evaluation of the FIVR system will be conducted at The Ohio State University to assess the usability and accuracy of food intake tracking with a group of 100 freeliving subjects, comparing 4 days of FIVR food intake data to 4 days of 24 hour recalls collected using ASA24 data. The resulting FIVR product will be a unique food intake tracker that combines selfadministration, automation (vision), and backend coding to collect food intake records to generate a detailed nutritional analysis.