a2 We proposed a conditional GAN approach to generate plausible videos from natural sounds. We used a 3D CNN model for video and CNN-RNN model for audio. Some Generated videos:
