A service has been created to determine the calorie content of food from a photo

Every person who at least once tried to go on a diet faced the need to count calories in their food. With this service, it is much easier to conduct such calculations.

service HuggingFace automatically recognizes the contents of the user’s plate and provides an approximate calorie content in it. The service is based on machine learning algorithms.

The user only needs to take a picture of the dish and upload its image to the service page. Algorithms compare photos with thousands of others in their database and display the result on the smartphone screen. The ratio of proteins, fats and carbohydrates is given, as well as the number of calories in the product.

The service is able to recognize the calorie content of not only individual products, but also ready meals, whether they are salads, hot dishes or desserts.

How to use the service

1. Click on the «Input» window that says «Drop Image Here or Click to Upload»
2. Taking a photo or uploading a previously taken product photo
3. Click «Submit»
4. We get the result in the «Output» window. It takes about 20 seconds to recognize a photo.

It is worth noting that the service does not always accurately recognize products. However, developers are constantly improving its algorithms. Probably in the future HuggingFace will be released as a separate mobile application. In the meantime, this is a pretty useful tool that should be bookmarked.