Enhancing Reward Models for High-quality Image Generation: Beyond Text-Image Alignment
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2025
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866912500996374528 |
|---|---|
| author | Ba, Ying Zhang, Tianyu Bai, Yalong Mo, Wenyi Liang, Tao Su, Bing Wen, Ji-Rong |
| author_facet | Ba, Ying Zhang, Tianyu Bai, Yalong Mo, Wenyi Liang, Tao Su, Bing Wen, Ji-Rong |
| contents | Contemporary image generation systems have achieved high fidelity and superior aesthetic quality beyond basic text-image alignment. However, existing evaluation frameworks have failed to evolve in parallel. This study reveals that human preference reward models fine-tuned based on CLIP and BLIP architectures have inherent flaws: they inappropriately assign low scores to images with rich details and high aesthetic value, creating a significant discrepancy with actual human aesthetic preferences. To address this issue, we design a novel evaluation score, ICT (Image-Contained-Text) score, that achieves and surpasses the objectives of text-image alignment by assessing the degree to which images represent textual content. Building upon this foundation, we further train an HP (High-Preference) score model using solely the image modality to enhance image aesthetics and detail quality while maintaining text-image alignment. Experiments demonstrate that the proposed evaluation model improves scoring accuracy by over 10\% compared to existing methods, and achieves significant results in optimizing state-of-the-art text-to-image models. This research provides theoretical and empirical support for evolving image generation technology toward higher-order human aesthetic preferences. Code is available at https://github.com/BarretBa/ICTHP. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2507_19002 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Enhancing Reward Models for High-quality Image Generation: Beyond Text-Image Alignment Ba, Ying Zhang, Tianyu Bai, Yalong Mo, Wenyi Liang, Tao Su, Bing Wen, Ji-Rong Computer Vision and Pattern Recognition Contemporary image generation systems have achieved high fidelity and superior aesthetic quality beyond basic text-image alignment. However, existing evaluation frameworks have failed to evolve in parallel. This study reveals that human preference reward models fine-tuned based on CLIP and BLIP architectures have inherent flaws: they inappropriately assign low scores to images with rich details and high aesthetic value, creating a significant discrepancy with actual human aesthetic preferences. To address this issue, we design a novel evaluation score, ICT (Image-Contained-Text) score, that achieves and surpasses the objectives of text-image alignment by assessing the degree to which images represent textual content. Building upon this foundation, we further train an HP (High-Preference) score model using solely the image modality to enhance image aesthetics and detail quality while maintaining text-image alignment. Experiments demonstrate that the proposed evaluation model improves scoring accuracy by over 10\% compared to existing methods, and achieves significant results in optimizing state-of-the-art text-to-image models. This research provides theoretical and empirical support for evolving image generation technology toward higher-order human aesthetic preferences. Code is available at https://github.com/BarretBa/ICTHP. |
| title | Enhancing Reward Models for High-quality Image Generation: Beyond Text-Image Alignment |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2507.19002 |