It contains a pytorch implementation of the following paper:
- MAU: A Motion-Aware Unit for Video Prediction and Beyond.
To install requirements:
pip install -r requirements.txt
To test our model on the Moving MNIST dataset, run the code in the PowerShell using:
python MAU_run.py --dataset mnist
To test our model on the Caltech Pedestrian dataset, run the code in the PowerShell using:
python MAU_run.py --dataset kitti
To test our model on the TownCentreXVID dataset, you should first download the sample data and the pretrained model and then place them in the following folders:
then run the code in the PowerShell using:
python MAU_run.py --dataset town
The predicted results will be saved into path results/.