Attention-Based Image-to-Video Translation for Synthesizing Facial Expression Using GAN

Figure 2

Channel attention architecture, adopted from [29].