G-Refine: A General Quality Refiner for Text-to-Image Generation
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , , , , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2024
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866910426644611072 |
|---|---|
| author | Li, Chunyi Wu, Haoning Hao, Hongkun Zhang, Zicheng Kou, Tengchaun Chen, Chaofeng Bai, Lei Liu, Xiaohong Lin, Weisi Zhai, Guangtao |
| author_facet | Li, Chunyi Wu, Haoning Hao, Hongkun Zhang, Zicheng Kou, Tengchaun Chen, Chaofeng Bai, Lei Liu, Xiaohong Lin, Weisi Zhai, Guangtao |
| contents | With the evolution of Text-to-Image (T2I) models, the quality defects of AI-Generated Images (AIGIs) pose a significant barrier to their widespread adoption. In terms of both perception and alignment, existing models cannot always guarantee high-quality results. To mitigate this limitation, we introduce G-Refine, a general image quality refiner designed to enhance low-quality images without compromising the integrity of high-quality ones. The model is composed of three interconnected modules: a perception quality indicator, an alignment quality indicator, and a general quality enhancement module. Based on the mechanisms of the Human Visual System (HVS) and syntax trees, the first two indicators can respectively identify the perception and alignment deficiencies, and the last module can apply targeted quality enhancement accordingly. Extensive experimentation reveals that when compared to alternative optimization methods, AIGIs after G-Refine outperform in 10+ quality metrics across 4 databases. This improvement significantly contributes to the practical application of contemporary T2I models, paving the way for their broader adoption. The code will be released on https://github.com/Q-Future/Q-Refine. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2404_18343 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | G-Refine: A General Quality Refiner for Text-to-Image Generation Li, Chunyi Wu, Haoning Hao, Hongkun Zhang, Zicheng Kou, Tengchaun Chen, Chaofeng Bai, Lei Liu, Xiaohong Lin, Weisi Zhai, Guangtao Multimedia Computer Vision and Pattern Recognition With the evolution of Text-to-Image (T2I) models, the quality defects of AI-Generated Images (AIGIs) pose a significant barrier to their widespread adoption. In terms of both perception and alignment, existing models cannot always guarantee high-quality results. To mitigate this limitation, we introduce G-Refine, a general image quality refiner designed to enhance low-quality images without compromising the integrity of high-quality ones. The model is composed of three interconnected modules: a perception quality indicator, an alignment quality indicator, and a general quality enhancement module. Based on the mechanisms of the Human Visual System (HVS) and syntax trees, the first two indicators can respectively identify the perception and alignment deficiencies, and the last module can apply targeted quality enhancement accordingly. Extensive experimentation reveals that when compared to alternative optimization methods, AIGIs after G-Refine outperform in 10+ quality metrics across 4 databases. This improvement significantly contributes to the practical application of contemporary T2I models, paving the way for their broader adoption. The code will be released on https://github.com/Q-Future/Q-Refine. |
| title | G-Refine: A General Quality Refiner for Text-to-Image Generation |
| topic | Multimedia Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2404.18343 |