CONTROLLABLE CONDITIONAL IMAGE GENERATION

A first generator stage of an image synthesis computer system generates an intermediate representation of an image that captures high-level structure of the image. The intermediate representation is generated from a random noise sample and without specific condition provided for the image. A second generator stage of the image synthesis computer system generates a conditional image based on i) the intermediate representation of the image and ii) a condition provided for the image, the condition specifying a desired style for the image. As a result, the conditional image is generated to match i) the high-level structure captured in the intermediate representation of the image and ii) the desired style specified by the condition provided for the image.