Scholarlink

Diabetes, particularly type 2, presents a significant health challenge in India, where traditional dietary practices complicate effective disease management, especially for elderly individuals. This paper introduces FoodWise, a comprehensive nutrition guide tailored to support people with diabetes in making informed dietary choices. FoodWise employs advanced technologies such as ResNet50 for food classification, Faster R-CNN for object detection, and the Segment Anything Model (SAM) for accurate food volume estimation. By incorporating fiducial markers, the system ensures precise portion measurements and integrates with the FatSecret API to provide detailed nutritional information, including calorie and nutrient content. The system demonstrated robust performance, achieving a top-1 accuracy of 93.43% in food recognition and reliable volume estimations. FoodWise offers a practical solution to address the dietary challenges faced by elderly individuals with diabetes, empowering them to manage their diet effectively and maintain better health outcomes.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

FoodWise: Diabetes-Friendly Nutrition Guide for Traditional Indian Dishes

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Radha Senthilkumar,
  • R. Srividya,
  • S. Abinaya,
  • B. A. Ahamed Alufar,
  • M. Premkumar,
  • P. Jayanthi

摘要

Diabetes, particularly type 2, presents a significant health challenge in India, where traditional dietary practices complicate effective disease management, especially for elderly individuals. This paper introduces FoodWise, a comprehensive nutrition guide tailored to support people with diabetes in making informed dietary choices. FoodWise employs advanced technologies such as ResNet50 for food classification, Faster R-CNN for object detection, and the Segment Anything Model (SAM) for accurate food volume estimation. By incorporating fiducial markers, the system ensures precise portion measurements and integrates with the FatSecret API to provide detailed nutritional information, including calorie and nutrient content. The system demonstrated robust performance, achieving a top-1 accuracy of 93.43% in food recognition and reliable volume estimations. FoodWise offers a practical solution to address the dietary challenges faced by elderly individuals with diabetes, empowering them to manage their diet effectively and maintain better health outcomes.