| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Personalized Image Generation | DreamBooth | CLIP-I Score90.6 | 34 | |
| Text-to-Image Personalization | DreamBooth original (test) | DINO Score0.771 | 18 | |
| Subject-driven image generation | DreamBooth Dataset 1.0 (test) | DINO Score0.774 | 18 | |
| Personalized Grounding | DreamBooth | Precision100 | 14 | |
| Object Compositing | DreamBooth (test) | Fidelity Score64.82 | 10 | |
| Image Editing | DreamBooth (evaluation set) | LPIPS0.322 | 8 | |
| Image Editing | DreamBooth (test) | SSIM74.4 | 8 | |
| Pairwise Object Compositing | DreamBooth (test) | FID255.5 | 8 | |
| Subject-driven generation | DreamBooth | Parameters (M)18.9 | 6 | |
| Subject-driven Image Generation | DreamBooth Average-case Generation | Subject Fidelity72.5 | 6 | |
| Subject-driven Image Generation | DreamBooth Best-case Generation | Subject Fidelity69.4 | 6 | |
| Object-driven generation | DreamBooth 30 classes | CLIP-T0.261 | 6 | |
| Identity-preserving Image Generation | DreamBooth (test) | Realism0.6284 | 6 | |
| Subject-driven generation | DreamBooth random 7 subjects | Subject Fidelity76.2 | 6 | |
| Subject-Style LoRA Fusion | DreamBooth | Style Similarity63 | 5 | |
| Furniture Synthesis | DreamBooth 15 | FID57.64 | 5 | |
| Subject-driven Generation | DreamBooth 25 subjects (test) | DINO0.703 | 5 | |
| Subject-Driven Text-to-Image Generation | DreamBooth | Preference Rate66.9 | 4 | |
| Subject-driven image generation | DreamBooth | Fine-tuning Loss0.086 | 4 | |
| Image Variation | DreamBooth and collected images (test) | Real-world Authenticity3.526 | 4 | |
| Subject-driven style generation | DreamBooth (test) | Subject Similarity0.462 | 3 | |
| Video Generation | DreamBooth | DINO Score61.4 | 3 | |
| Subject-driven Text-to-Image Generation | DreamBooth Stable Diffusion v1.4 (test) | CLIP-I77.4 | 3 | |
| Image Editing | DreamBooth | CLIP-T28.69 | 3 | |
| Subject-driven Image Generation | DreamBooth SDXL (val) | Win Rate45.1 | 2 |