Research map / Image generation / Text-to-image diffusion
Energy-Based Cross Attention for Bayesian Context Update in Text-to-Image Diffusion Models
Geon Yeong Park, Jeongsol Kim, Beomsu Kim, Sang Wan Lee, Jong Chul Ye
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Use ours instead of pix2pix-zero | image editing · Cat -> Dog | CLIP accuracy: 92.4% -> 93.7% | Improved |
| Use ours instead of pix2pix-zero | image editing · Cat -> Dog | structure distance: 0.044 -> 0.040 | Improved |
| Use ours instead of pix2pix-zero | image editing · Horse -> Zebra | CLIP accuracy: 75.2% -> 90.4% | Improved |
| Use ours instead of pix2pix-zero | image editing · Horse -> Zebra | structure distance: 0.066 -> 0.061 | Improved |
| Use ours instead of pix2pix-zero | image editing · Cat -> Cat w/ glasses | CLIP accuracy: 71.2% -> 81.1% | Improved |
| Use ours instead of pix2pix-zero | image editing · Cat -> Cat w/ glasses | structure distance: 0.028 -> 0.052 | Hurt |
| Use ours instead of DDIM + word swap | image editing · Horse -> Zebra | CLIP accuracy: 94.0% -> 90.4% | Hurt |
| Use ours instead of DDIM + word swap | image editing · Horse -> Zebra | structure distance: 0.123 -> 0.061 | Improved |
| Use ours instead of prompt-to-prompt | image editing · Horse -> Zebra | CLIP accuracy: 18.4% -> 90.4% | Improved |
| Use ours instead of SDEdit + word swap | image editing · Cat -> Dog | CLIP accuracy: 71.2% -> 93.7% | Improved |
| Increase gamma_attn (attention step size) | EBCU · multi-concept generation | image quality (qualitative) | Hurt |
| Increase gamma_reg (regularisation step size) | EBCU · multi-concept generation | image quality (qualitative) | Hurt |
Closest papers by meaning
- Cross-Attention Head Position Patterns Can Align with Human Visual Concepts in Text-to-Image Generative Models (ICLR 2025 · similarity 0.88)
- R&B: Region and Boundary Aware Zero-shot Grounded Text-to-image Generation (ICLR 2024 · similarity 0.88)
- Energy-Guided Optimization for Personalized Image Editing with Pretrained Text-to-Image Diffusion Models (AAAI 2025 · similarity 0.88)
- Noise Diffusion for Enhancing Semantic Faithfulness in Text-to-Image Synthesis (CVPR 2025 · similarity 0.88)
- OneActor: Consistent Subject Generation via Cluster-Conditioned Guidance (NeurIPS 2024 · similarity 0.87)
- AP-Adapter: Improving Generalization of Automatic Prompts on Unseen Text-to-Image Diffusion Models (NeurIPS 2024 · similarity 0.87)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
