A Model-Agnostic Semantic-Quality Compatible Framework based on Self-Supervised Semantic Decoupling

Xiaoyu Ma,Chenxi Feng,Jiaojiao Wang,Qiang Lin,Suiyu Zhang,Jinchi Zhu,Xiaodiao Chen,Chang Liu,Dingguo Yu
DOI: https://doi.org/10.1145/3581783.3613775
2023-01-01
Abstract:Blind Image Quality Assessment (BIQA) is a challenging research topic that is critical for preprocessing and optimizing downstream vision tasks such as semantic recognition and image restoration. However, there has been a significant disconnect between BIQA research and other vision tasks. The primary cause of such disconnect is the incompatibility of existing BIQA models with other vision tasks, resulting in significant computational complexity. To address this issue, we propose a model-agnostic semantic-quality compatible framework that can simultaneously generate quality and semantic predictions. By incorporating a lightweight learning architecture, we demonstrate that a parameter-fixed semantic-oriented backbone can predict the perceptual quality of images as accurately as models trained end-to-end. We systematically study the major components of our framework, and our experimental results demonstrate the superiority of our model in terms of both complexity and accuracy. The source code of this work is available at https://github.com/MaxiaoyuHehe/SQCFNet.
What problem does this paper attempt to address?