About UFall

UFall performs first-person fall detection using mobile devices like a phone or table or wearable IP camera device using Computer Vision and Machine Learning Techniques. In particular, an LSTM model is developed to recognize the motion involved in a first-person video during a fall event. This tool is currently developed using Tensorflow framework and python. Once LSTM is supported in Tensorflow Lite (Mobile) then plans for deployment to run on intially Android devices are in place. Currently, the mobile device acts like an IP camera and transmits video to a machine running Tensorflow in Python where LSTM is supported.