Table 1. Training and regressor comparison.

Category	Method	PLCC ↑	SRCC ↑
Optimizer	SGD with momentum	0.852	0.840
	AdamW (ours)	0.912	0.901
Loss function	MSE	0.881	0.869
	MSE+Ranking loss (ours)	0.912	0.901
Regressor	Linear regressor	0.864	0.851
	MLP regressor (ours)	0.912	0.901