Scene Graph Conditioning in Latent Diffusion
This work addresses the challenge of fine-grained control in image generation for applications requiring precise scene composition, though it is incremental as it builds on existing techniques like ControlNet and Gated Self-Attention.
The paper tackles the problem of achieving detailed semantic control in diffusion-based image generation by using scene graphs instead of ambiguous text prompts, resulting in higher quality image synthesis that outperforms previous methods.
Diffusion models excel in image generation but lack detailed semantic control using text prompts. Additional techniques have been developed to address this limitation. However, conditioning diffusion models solely on text-based descriptions is challenging due to ambiguity and lack of structure. In contrast, scene graphs offer a more precise representation of image content, making them superior for fine-grained control and accurate synthesis in image generation models. The amount of image and scene-graph data is sparse, which makes fine-tuning large diffusion models challenging. We propose multiple approaches to tackle this problem using ControlNet and Gated Self-Attention. We were able to show that using out proposed methods it is possible to generate images from scene graphs with much higher quality, outperforming previous methods. Our source code is publicly available on https://github.com/FrankFundel/SGCond