Closed manmanCover closed 5 years ago
According to paper DELVING DEEPER INTO CONVOLUTIONAL NETWORKS FOR LEARNING VIDEO REPRESENTATIONS (https://arxiv.org/abs/1511.06432), there are 6 2D-convolutions. However, in your implementation, there are only 3, why?
According to paper DELVING DEEPER INTO CONVOLUTIONAL NETWORKS FOR LEARNING VIDEO REPRESENTATIONS (https://arxiv.org/abs/1511.06432), there are 6 2D-convolutions. However, in your implementation, there are only 3, why?