Text-to-image generation on Landmarks
25.329CLIPScore-TTextTIGER
Evaluation results
| Method | Links | ||
|---|---|---|---|
| TextTIGERBase Model=SD3.5, LLM Backbone=GPT-52025.04 | 25.329 | 24.729 | |
| TextTIGERBase Model=Dreamlike, LLM Backbone=GPT-52025.04 | 24.829 | 24.822 | |
| TextTIGERBase Model=Qwen-Img, LLM Backbone=GPT-52025.04 | 24.414 | 24.615 | |
| DreamlikeMode=Cap-Only2025.04 | 23.978 | 20.556 | |
| SD3.5Mode=Cap-Only2025.04 | 23.882 | 20.532 | |
| TextTIGERBase Model=FLUX, LLM Backbone=GPT-52025.04 | 23.669 | 25.169 | |
| TextTIGERBase Model=PixArt, LLM Backbone=GPT-52025.04 | 23.144 | 24.939 | |
| Qwen-ImgMode=Cap-Only2025.04 | 22.417 | 20.614 | |
| FLUXMode=Cap-Only2025.04 | 21.454 | 20.837 | |
| RAGBase Model=Dreamlike, Strategy=Aug-Only2025.04 | 20.939 | 19.092 | |
| RAGBase Model=Dreamlike, Strategy=BM252025.04 | 20.743 | 19.934 | |
| PixArtMode=Cap-Only2025.04 | 19.106 | 20.491 |