P3S-Diffusion:A Selective Subject-driven Generation Framework via Point Supervision
This work addresses a domain-specific problem in image generation for users needing precise subject selection, offering an incremental improvement over existing methods by reducing reliance on expensive masks or imprecise text prompts.
The paper tackles the challenge of accurately selecting similar subjects in images for subject-driven generation by introducing P3S-Diffusion, a framework that uses point supervision to generate subject-driven images with minimal labeling cost, achieving excellent feature preservation and generation capabilities as demonstrated in experiments.
Recent research in subject-driven generation increasingly emphasizes the importance of selective subject features. Nevertheless, accurately selecting the content in a given reference image still poses challenges, especially when selecting the similar subjects in an image (e.g., two different dogs). Some methods attempt to use text prompts or pixel masks to isolate specific elements. However, text prompts often fall short in precisely describing specific content, and pixel masks are often expensive. To address this, we introduce P3S-Diffusion, a novel architecture designed for context-selected subject-driven generation via point supervision. P3S-Diffusion leverages minimal cost label (e.g., points) to generate subject-driven images. During fine-tuning, it can generate an expanded base mask from these points, obviating the need for additional segmentation models. The mask is employed for inpainting and aligning with subject representation. The P3S-Diffusion preserves fine features of the subjects through Multi-layers Condition Injection. Enhanced by the Attention Consistency Loss for improved training, extensive experiments demonstrate its excellent feature preservation and image generation capabilities.