AI tools has not been easy to use in specific design mission. I tried some new process, the effect of specific generation is very poor, but it is still useful as an auxiliary tool to speed up. The character "疾", a co-creation IP made with @J表面. The process is roughly sketch-dazpose (used to produce depth and openpose bones), SD produces a large number of pictures with the same pose, and then returns to PS to stitch and refine the pictures. Relatively slow, but more controllable, AI is mainly used for detail divergence and filling of non-important details