MIT Researchers have developed deep-learning algorithm that can compile a list of ingredients and even recommend recipes after looking at photos of food.
Their AI system named Pic2Recipe predicts a dish’s recipe from just a photo.
Pic2Recipe reads through over one million recipes. Then used a neural network to find patterns and connections between images and recipes.
Researchers train a joint embedding composed of an encoder for each modality (ingredients, instructions and images) and evaluate all the recipe representations for im2recipe retrieval. Given a food image, the task is to retrieve its recipe from a collection of test recipes.
News Source: http://im2recipe.csail.mit.edu/
See the demo of Pic2Recipe at http://tuesday.csail.mit.edu:4242/
