This repository is the Website of the FoodScore project developed for food recognition using deep learning.
You can have a look at the Website, but unfortunately the prediction can only be done once you train the model and get your own weights and docker image, as this comes with associated costs.
Therefore, it is recommended to read first that repository in order to understand how the model works and how you can get your weights.
Feel free to ⭐ and clone this repo 😉
The Website is very easy to use as this is important for every business solution. First, you need to upload an image that contains at least one food category of the Food100 dataset.
Then you can move the white square on the left image to select the food item. You can see the cropped image in the right side.
Once you have selected the food item, click on the show nutritions
to get the prediction and the nutrients.