|
Briefing
|
- Pic2Recipe – MIT Computer Science and Artificial Intelligence Laboratory (CSAIL) researchers developed Pic2Recipe, artificial intelligence algorithm capable of naming food in images, identifying ingredients and suggesting similar recipes
- Neural Network Training – Trained using Recipe1M, database of one million food images from food-related websites, such as All Recipes and Food.com, with corresponding information on recipes and ingredients
- Potential Applications – Could be used to understand eating habits, monitor daily nutrition, and create home versions of food found in restaurants
- Organizations Involved – Co-developed by researchers from Polytechnic University of Catalonia and Qatar Computing Research Institute (QCRI), and funded by QCRI, European Regional Development Fund, and the Spanish Ministry of Economy, Industry, and Competitiveness
- Next Steps – Include expansion of dataset, addition of more metrics, such as volume and healthiness, and enhancements of Pic2Recipe's capability to understand food in detail, including ability to determine how food is prepared
|
|
Organization
|
European Regional Development Fund (ERDF), MIT Computer Science and Artificial Intelligence Laboratory (CSAIL), Qatar Computing Research Institute, Spanish Ministry of Economy Industry and Competitiveness
|
