foodE Helper

To begin, users can import a food picture from the Camera or Photo Library to be analyzed. Using an image-recognition machine-learning model trained on tens of thousands of input pictures, foodE Helper categorizes the food in the provided picture as one of over 100 different types of dishes.

foodE Helper then makes calls to the Food2Fork API to determine the ingredients used to prepare the dish. By processing the data returned by the API, foodE Helper identifies the possible presence of dozens of food products relevant to a variety of dietary restrictions.

By default, foodE Helper informs users of the possible presence of all food products in its database. In foodE Helper’s settings, however, users have the option of selecting only the food products relevant to their dietary restrictions - for example, allergens or ingredients violating special diets. Based on these user selections, foodE Helper will adjust its analyses to display only the food products relevant to a particular user.

Important note: the data on which foodE Helpers image-recognition machine-learning model was trained may not encompass every possible dish relevant to any particular user; the model itself may occasionally misclassify dishes, as its accuracy is not perfect (100%). In addition, foodE Helpers database may not encompass every possible food product relevant to any particular user. Predictions made by foodE Helper are conjectural and should be confirmed by the user, especially when high risk is involved (such as in the case of severe allergens).