Implementation of Real-world Anomaly Detection in Surveillance Videos paper from CVPR 2018. Original implementation from authors is here.
Download C3D sports-1m weights from here and save them to 'trained_models' folder as 'c3d_sports1m.h5'.
Run demo.py to run the code on the demo video in 'input' folder. Visualization of the predictions from the model is saved in 'output' folder.
- Keras : 2.2.0
- Tensorflow : 1.10.1
- Numpy : 1.14.5
- OpenCV : 3.3.0.10
- Scipy : 0.19.1
- Matplotlib : 2.0.2
Will add the training code soon...!!