UniPercept: Towards Unified Perceptual-Level Image Understanding across Aesthetics, Quality, Structure, and Texture
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866908731953905664 |
|---|---|
| author | Cao, Shuo Li, Jiayang Li, Xiaohui Pu, Yuandong Zhu, Kaiwen Gao, Yuanting Luo, Siqi Xin, Yi Qin, Qi Zhou, Yu Chen, Xiangyu Zhang, Wenlong Fu, Bin Qiao, Yu Liu, Yihao |
| author_facet | Cao, Shuo Li, Jiayang Li, Xiaohui Pu, Yuandong Zhu, Kaiwen Gao, Yuanting Luo, Siqi Xin, Yi Qin, Qi Zhou, Yu Chen, Xiangyu Zhang, Wenlong Fu, Bin Qiao, Yu Liu, Yihao |
| contents | Multimodal large language models (MLLMs) have achieved remarkable progress in visual understanding tasks such as visual grounding, segmentation, and captioning. However, their ability to perceive perceptual-level image features remains limited. In this work, we present UniPercept-Bench, a unified framework for perceptual-level image understanding across three key domains: Aesthetics, Quality, Structure and Texture. We establish a hierarchical definition system and construct large-scale datasets to evaluate perceptual-level image understanding. Based on this foundation, we develop a strong baseline UniPercept trained via Domain-Adaptive Pre-Training and Task-Aligned RL, enabling robust generalization across both Visual Rating (VR) and Visual Question Answering (VQA) tasks. UniPercept outperforms existing MLLMs on perceptual-level image understanding and can serve as a plug-and-play reward model for text-to-image generation. This work defines Perceptual-Level Image Understanding in the era of MLLMs and, through the introduction of a comprehensive benchmark together with a strong baseline, provides a solid foundation for advancing perceptual-level multimodal image understanding. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2512_21675 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | UniPercept: Towards Unified Perceptual-Level Image Understanding across Aesthetics, Quality, Structure, and Texture Cao, Shuo Li, Jiayang Li, Xiaohui Pu, Yuandong Zhu, Kaiwen Gao, Yuanting Luo, Siqi Xin, Yi Qin, Qi Zhou, Yu Chen, Xiangyu Zhang, Wenlong Fu, Bin Qiao, Yu Liu, Yihao Computer Vision and Pattern Recognition Multimodal large language models (MLLMs) have achieved remarkable progress in visual understanding tasks such as visual grounding, segmentation, and captioning. However, their ability to perceive perceptual-level image features remains limited. In this work, we present UniPercept-Bench, a unified framework for perceptual-level image understanding across three key domains: Aesthetics, Quality, Structure and Texture. We establish a hierarchical definition system and construct large-scale datasets to evaluate perceptual-level image understanding. Based on this foundation, we develop a strong baseline UniPercept trained via Domain-Adaptive Pre-Training and Task-Aligned RL, enabling robust generalization across both Visual Rating (VR) and Visual Question Answering (VQA) tasks. UniPercept outperforms existing MLLMs on perceptual-level image understanding and can serve as a plug-and-play reward model for text-to-image generation. This work defines Perceptual-Level Image Understanding in the era of MLLMs and, through the introduction of a comprehensive benchmark together with a strong baseline, provides a solid foundation for advancing perceptual-level multimodal image understanding. |
| title | UniPercept: Towards Unified Perceptual-Level Image Understanding across Aesthetics, Quality, Structure, and Texture |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2512.21675 |