AnomalyDetection-in-SurveillanceVideos

Implementation of Convolutional Autoencoders(spatio-temporal) to be trained over Avenue and UCF Datasets to identify anomalous activities.

How to Run the Code

Load all the '.py' files in Spyder, then you first run Preprocessing file that will extract the frames from the Avenue and UCSD Datasets and follow the instructions in the file that are commented. Then Run Model_Wrapper.py and move to the training of the model with Train.py after training is completed you can then Deploy the model with Develop_Model.py. All the files are commented and documented and the code is self explanatory so I hope there won't be any issues.

Training Acc & Training Loss