UniPercept: Towards Unified Perceptual-Level Image Understanding across Aesthetics, Quality, Structure, and Texture

TL;DR

UniPercept introduces a unified framework and benchmark for perceptual-level image understanding across aesthetics, quality, structure, and texture, enhancing multimodal models' perceptual capabilities.

Contribution

This work establishes a hierarchical definition system, large-scale datasets, and a strong baseline for perceptual-level image understanding in multimodal models.

Findings

UniPercept outperforms existing models on perceptual understanding tasks.

It enables robust generalization across VR and VQA tasks.

Can serve as a plug-and-play reward model for text-to-image generation.

Abstract

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.