AttrLostGAN: Attribute Controlled Image Synthesis from Reconfigurable Layout and Style
This work addresses the need for user-friendly interfaces in image generation applications by enabling attribute-level control without affecting other parts of the image, representing an incremental improvement over existing layout-to-image methods.
The paper tackles the problem of fine-grained control over individual object appearances in conditional image synthesis from layout, achieving successful control of object details in complex multi-object scenes.
Conditional image synthesis from layout has recently attracted much interest. Previous approaches condition the generator on object locations as well as class labels but lack fine-grained control over the diverse appearance aspects of individual objects. Gaining control over the image generation process is fundamental to build practical applications with a user-friendly interface. In this paper, we propose a method for attribute controlled image synthesis from layout which allows to specify the appearance of individual objects without affecting the rest of the image. We extend a state-of-the-art approach for layout-to-image generation to additionally condition individual objects on attributes. We create and experiment on a synthetic, as well as the challenging Visual Genome dataset. Our qualitative and quantitative results show that our method can successfully control the fine-grained details of individual objects when modelling complex scenes with multiple objects. Source code, dataset and pre-trained models are publicly available (https://github.com/stanifrolov/AttrLostGAN).