Students' Projects

Human Activity Recognition using Event based Sensor for Home Application
Department Information Technology
Year 2019-20
Budget 0
Students Priyal Rajeshbhai Patel , Shruti Rajeshkumar Patel, Bhavik Pankajbhai Parmar, Dishang Vikasbhai Patel, Parth Vallabhbhai Pansuriya, Nisarg Devesh Chokshi
Mentor Prof. Safvan Vahora
Abstract As per our proposal, we use DVS datasets as an input video, apply some necessary operations on it and trained our model and tested it using the datasets, and get the Activity prediction based on the filtering mechanism of the model. Based on our model, we classify various activities and use them for home IOT applications. Since we are thinking of for home IoT appliances which must operate with low power consumption and low latency, this needs fairly match with DVS sensor-based activity recognition that provides fast processing, low power consumption, low latency and less storage requirement.