visionMaze

we apply the masks on the original pixels. The masks are used to only keep the subject of the images.

Thank you for your interest. Our framework is suitable for most LDM models. The integration efforts differ based on the compatibility of open-sourced implementation. Our near-term plan is to integrate...