How often do we not struggle to decide on what to cook for dinner? Or see some mouthwatering dishes on the television/ internet but don’t know what it’s called? Or that situation we experience when we see something delicious and exciting on the adjacent table? Or unable to find the local name for a food item in a foreign country/ foreign language?
Food Peek helps you tackle such exact situations. Food Peek makes use of efficient machine learning algorithms to successfully classify the image input by the user and output the name, ingredients, recipes, nutrition content and sites from where the item can be ordered.
FRONT-END: TKINTER MODULE
BACK-END: MYSQL DATABASE SERVER
DATASET USED CAN BE FOUND ON KAGGLE: Food Images - Food 101
- Download the Python3 notebook and open it in your choice of editor (Jupyter Notebook, Jupyter Lab, etc.)
- Download the Food Image dataset, test images and images used in the GUI.
- Change the path for the loction of your food image dataset and also for the test images.
- Change the path of the images used in the GUI to the current path where it is stored.
- Update your EMAIL ID and PASSWORD under the email section of the application.
- Update your USERNAME and PASSWORD to connect to the MySQL Database server.
- Run each cell and train the ML model. Once the model is trained and ready, you will be able to use it in your application.
- You should have pre-installed MySQL server on your PC to be able to connect and store the incoming data via the application. If you don't have it installed, then do so and keep it ready.
- Inorder to send emails via Google, you need to TURN ON -LESS SECURE APP ACCESS to be able to login using the Python code. Don't forget to turn this setting OFF once your are done using the application in order to maintain the security of your Google Account.