FYP - Final Year Project - B.Eng in Mechatronics Engineering: Calorie-counting/diet-management Android application using ML image classifier and Bluetooth kitchen scale. This app uses the TensorFlow machine learning library, trained with the MobileNet image classification model, to detect food types. With the food classified, nutritional info is retrieved from the USDA nutritional info database, and the user can use a Bluetooth kitchen scale to record the weight value. This is used to calculate the calories and macronutrients of the food item or overall meal. Can be used with text-input for search and weight values also. Currently integrated with the "Skale 2" Bluetooth scale, but further support being added. Work in progress.

Java

Final Year Project - B.Eng in Mechatronic Engineering

This Android app was built for the Final Year Project of my B.Eng in Dublin City University, Ireland. It was an exploration into the capabilities for ML-based image classification to innovate within the calorie-counting/diet-management applicaton industry. Namely, the application uses a retrained version of the MobileNets image classification model to perform this on Android - implemented with the lightweight TensorFlow Lite ML library. This was on the leading edge of mobile-based ML in 2017/18!

Please excuse the code quality: my project was based on the apps' functionality and on my research, so code quality lapsed in the rush to complete the project. It will be cleaned up eventually!

LukeScales1/FYP

Final Year Project - B.Eng in Mechatronic Engineering