TABLE 2.

Performance comparison on CVIQD database.

Models	JPEG			AVC			HEVC			ALL
	PLCC	SRCC	RMSE	PLCC	SRCC	RMSE	PLCC	SRCC	RMSE	PLCC	SRCC	RMSE
BRISQUE (Mittal et al., 2012)	0.9519	0.9308	4.9825	0.8913	0.8559	5.6647	0.8979	0.8980	5.3367	0.9001	0.8814	6.2327
DIIVINE (Moorthy and Bovik, 2011)	0.9331	0.8710	6.0475	0.9024	0.8927	5.0618	0.9031	0.8530	5.6397	0.8988	0.9080	6.0260
SSEQ (Liu et al., 2014)	0.9745	0.9527	4.0731	0.9381	0.9180	4.2228	0.9115	0.9059	5.0884	0.9263	0.9134	5.2609
OG-IQA (Liu et al., 2016)	0.9745	0.9261	3.6130	0.8871	0.8852	5.7588	0.9030	0.9055	4.6374	0.9197	0.8969	5.3562
NRSL (Li et al., 2016)	0.9570	0.9056	5.1460	0.9145	0.8823	4.9565	0.9000	0.8981	4.8063	0.8850	0.8944	6.8612
SSP-BOIQA (Zheng et al., 2020)	0.915	0.853	6.847	0.885	0.861	7.042	0.854	0.841	6.302	0.890	0.856	6.941
MC360IQA (Sun et al., 2019)	0.9410	0.9230	5.8040	0.9320	0.9410	5.3570	0.9140	0.8990	4.8010	0.9390	0.9040	4.6060
Zhou Y. et al. (2022)	0.957	0.923	5.601	0.953	0.949	3.873	0.929	0.914	4.525	0.902	0.911	6.117
S3DAVS	0.9707	0.9302	3.8675	0.9586	0.9447	3.3925	0.9367	0.8802	4.5675	0.9533	0.9426	4.1022

The best-performing NR metrics are highlighted in bold.