Sensing, Recognition, and Security / Touchscreen Input Modeling and Generation

How can GANs simultaneously model spatial distribution and temporal information of touchscreen input?