Abstract: The current state-of-the-art No-Reference Image Quality Assessment (NR-IQA) methods typically use feature extraction in upstream backbone networks, which assumes that all extracted features are relevant. However, we argue that not all features are beneficial, and some may even be harmful, necessitating careful selection. Empirically, we find that many image pairs with small feature spatial distances can have vastly different quality scores. To address this issue, we propose a Quality-Aware Feature Matching IQA metric~(QFM-IQM) that employs contrastive learning to remove harmful features from the upstream task. Specifically, our approach enhances the semantic noise distinguish capabilities of neural networks by comparing image pairs with similar quality scores but varying semantic features and adaptively adjusting the upstream task’s features by introducing disturbance. Furthermore, we utilize a distillation framework to expand the dataset and improve the model's generalization ability. Our approach achieves superior performance to the state-of-the-art NR-IQA methods on 8 standard NR-IQA datasets, achieving PLCC values of \textbf{0.932} (\textcolor{red}{$\uparrow 2.4\%$} vs. 0.908 in TID2013) and \textbf{0.913} (\textcolor{red}{$\uparrow 1.9\%$} vs. 0.894 in LIVEC).