Table 5 Comparison of no-reference image quality metrics for various enhancement techniques on the LIME dataset³.

Metric	Paired supervision				Unsupervised	Zero-shot learning
Metric	BIMEF¹⁰²	LIME³	MF¹⁰³	Multiscale Retinex¹⁰⁴	EnlightenGAN⁵⁶	SG-ZSL⁵⁴	Zero-DCE⁵²	Zero-DCE++⁵³	Ours
NIMA⁹¹	4.48	4.48	4.53	4.36	4.47	4.47	4.37	4.39	4.51
PaQ2PiQ⁹²	73.51	75.82	74.83	74.32	74.58	74.58	74.94	74.96	74.72
DBCNN⁹³	52.13	48.90	49.41	51.06	48.57	48.57	46.81	46.83	52.51
musiq-koniq⁵⁵	61.29	61.25	61.58	61.58	60.35	60.35	59.59	59.61	61.03
MANIQA⁹⁴	0.64	0.63	0.64	0.58	0.61	0.61	0.59	0.61	0.61
CLIPIQA⁹⁵	0.42	0.42	0.42	0.35	0.40	0.40	0.33	0.35	0.40
TReS-koniq⁹⁶	69.50	63.33	66.35	66.98	66.08	66.79	67.77	67.79	69.69
HyperIQA⁹⁷	0.51	0.48	0.49	0.49	0.50	0.50	0.48	0.50	0.52
GPR-BIQA¹⁰	0.63	0.63	0.64	0.62	0.63	0.63	0.61	0.63	0.68
Quality Net⁹⁸	0.64	0.65	0.66	0.64	0.66	0.66	0.63	0.64	0.69
PIQI⁹	0.63	0.64	0.65	0.63	0.65	0.65	0.62	0.61	0.67
Average	24.03	23.38	23.65	23.78	23.41	23.47	23.34	23.36	24.18

Quick links

Search