##Abstract LRCN is a combination of CNN and RNN
##Installation
- Download UCF101 dataset
- ./avi2jpg.sh to split the video of UCF101
- Download the pretrained model of vgg16
- python cnn_predict.py to get the pretrained data
- python train_ucf101.py
##Reference Donahue, J., Hendricks, L. A., Guadarrama, S., Rohrbach, M., Venugopalan, S., Saenko, K., & Darrell, T. (2014). Long-term recurrent convolutional networks for visual recognition and description. arXiv preprint arXiv:1411.4389.Chicago