Q-Ponder：面向基于推理的视觉质量评估的统一训练流水线
Q-Ponder: A Unified Training Pipeline for Reasoning-based Visual Quality Assessment
AI总结：
针对现有视觉质量评估方法中打分精度与推理可解释性难以兼顾的问题，提出两阶段统一训练框架Q-Ponder，结合知识蒸馏与带新型奖励的GRPO强化学习，实现了打分性能与描述质量的双重SOTA提升。
摘要：
近期研究表明，多模态大语言模型（MLLMs）能够通过可解释评估熟练完成视觉质量评价。然而现有方法通常将质量打分与推理描述视为相互独立的任务，采用互不关联的优化目标，由此产生权衡问题：擅长质量推理描述的模型难以实现精准的分数回归，而专注于打分的模型又缺乏可解释性。这一局限限制了MLLMs在视觉质量评估领域的全部潜力，而该领域中准确性与可解释性本应相辅相成。为解决这一问题，我们提出了一个包含冷启动阶段和基于强化学习的微调阶段的两阶段统一训练框架。具体而言，第一阶段通过专家设计的提示词从教师模型中蒸馏高质量数据，借助交叉熵损失监督初始化推理能力。第二阶段我们引入一种新型奖励机制，结合Group Relative Policy Optimization（GRPO，组相对策略优化）联合优化打分准确性与推理一致性。我们将这两个阶段得到的模型分别命名为Q-Ponder-CI和Q-Ponder。大量实验表明，Q-Ponder在质量分数回归基准上达到了当前最优（SOTA）性能，在跨域数据集上的SRCC最高提升6.5%。此外，Q-Ponder显著优于基于描述的SOTA模型，包括其教师模型Qwen-2.5-VL-72B，尤其在描述准确性和合理性方面表现突出，展现了在多样化任务上的泛化潜力。
英文摘要：
Recent studies demonstrate that multimodal large language models (MLLMs) can proficiently evaluate visual quality through interpretable assessments. However, existing approaches typically treat quality scoring and reasoning descriptions as separate tasks with disjoint optimization objectives, leading to a trade-off: models adept at quality reasoning descriptions struggle with precise score regression, while score-focused models lack interpretability. This limitation hinders the full potential of MLLMs in visual quality assessment, where accuracy and interpretability should be mutually reinforcing. To address this, we propose a unified two-stage training framework comprising a cold-start stage and a reinforcement learning-based fine-tuning stage. Specifically, in the first stage, we distill high-quality data from a teacher model through expert-designed prompts, initializing reasoning capabilities via cross-entropy loss supervision. In the second stage, we introduce a novel reward with Group Relative Policy Optimization (GRPO) to jointly optimize scoring accuracy and reasoning consistency. We designate the models derived from these two stages as Q-Ponder-CI and Q-Ponder. Extensive experiments show that Q-Ponder achieves state-of-the-art (SOTA) performance on quality score regression benchmarks, delivering up to 6.5% higher SRCC on cross-domain datasets. Furthermore, Q-Ponder significantly outperforms description-based SOTA models, including its teacher model Qwen-2.5-VL-72B, particularly in description accuracy and reasonableness, demonstrating the generalization potential over diverse tasks.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
