SCS Undergraduate Thesis Topics
|Geeta Shroff||Asim Smailagic||DiaWear: An Assistive Wearable Food and Activity Recognition, Monitoring and Management System for Diabetes Patients|
DiaWear is a wearable food and activity recognition system to monitor and aid medical patients with respect to calorie intake and consumption. Such a system will be ideal for diabetes patients, as well as pancreatitis, and other patients with dietary restrictions. This system may also find use in fitness and weight management. Gathering data from sensors placed on different parts of the body using existing platforms such as the CMU e-Watch and the CMU ArmBand, and by using a PDA equipped with accelerometers, information regarding the activity and exercise of the user will be recorded and analysed via machine learning algorithms. A food recognition algorithm on the PDA will allow the system to take pictures and recognize certain learned foods on a plate and relay the corresponding calories. Knowledge of calories burnt via merged sensor data and knowledge of available calories via food recognition, will allow the system to monitor and assist the user regarding calorie intake a! nd management on a regular basis. Recorded data will also help medical professionals to provide more accurate forms of treatment and medications. As part of a first semester CS Senior Thesis project, the primary role of the student will be to develop an image recognition algorithm to detect food, and future work, if time permits, will comprise of integrating this algorithm into the above mentioned system of wearable sensors.