I have to develop a video event detection tool in ticketing counter.The tool must take photographs of persons who jumb over gate without taking tickets.I have a set of videos in which people jumbing over gate.So with that data in hand How can i implement a Video event detection tool?I am new to video analytics.I don't know where to start.Where can i find some good tutorial about supervised video event detection.?
I recommend using a 3D convoluted neural network. Traditional 2D convoluted neural networks excel at identifying objects in images, and 3D convolution is used for identifying objects or movements also across multiple frames (i.e., time). Tensorflow offers this capability.