Paper Detecting anomalous events in videos by learning deep representations of appearance and motion on python, opencv and tensorflow. This paper uses the stacked denoising autoencoder for the the feature training on the appearance and motion flow features as input for different window size and using multiple SVM as a single c

DeepvisionAI-backend is the code core of a desktop software providing detection/tracking of pedestrians and abnormal crowd activity using convolutional auto-encoders.

Events study: Kalman filter and neural network applications with a case study on M&A price run-ups