DPC-VQA: Decoupling Quality Perception and Residual Calibration for Video Quality Assessment

TL;DR

DPC-VQA introduces a decoupling framework that leverages pretrained multimodal large language models for video quality assessment, reducing training costs while maintaining competitive performance.

Contribution

It proposes a lightweight calibration approach that adapts a frozen MLLM's perceptual prior to new scenarios without extensive retraining.

Findings

Achieves competitive performance with less than 2% trainable parameters.

Remains effective with only 20% of MOS labels.

Reduces training and data costs significantly.

Abstract

Recent multimodal large language models (MLLMs) have shown promising performance on video quality assessment (VQA) tasks. However, adapting them to new scenarios remains expensive due to large-scale retraining and costly mean opinion score (MOS) annotations. In this paper, we argue that a pretrained MLLM already provides a useful perceptual prior for VQA, and that the main challenge is to efficiently calibrate this prior to the target MOS space. Based on this insight, we propose DPC-VQA, a decoupling perception and calibration framework for video quality assessment. Specifically, DPC-VQA uses a frozen MLLM to provide a base quality estimate and perceptual prior, and employs a lightweight calibration branch to predict a residual correction for target-scenario adaptation. This design avoids costly end-to-end retraining while maintaining reliable performance with lower training and data costs. Extensive experiments on both user-generated content (UGC) and AI-generated content (AIGC) benchmarks show that DPC-VQA achieves competitive performance against representative baselines, while using less than 2% of the trainable parameters of conventional MLLM-based VQA methods and remaining effective with only 20\% of MOS labels. The code will be released upon publication.