Table 4 Comparison of no-reference image quality metrics for various enhancement techniques on the DICM dataset⁷⁸.

Metric	Paired supervision				Unsupervised	Zero-shot learning
Metric	BIMEF¹⁰²	LIME³	MF¹⁰³	Multiscale Retinex¹⁰⁴	EnlightenGAN⁵⁶	SemZSL⁵⁴	Zero-DCE⁵²	Zero-DCE++⁵³	Ours
NIMA⁹¹	4.52	4.42	4.56	4.58	4.29	4.29	4.22	4.24	4.36
PaQ2PiQ⁹²	74.58	76.17	75.60	74.42	73.37	73.31	73.91	73.93	76.95
DBCNN⁹³	51.36	50.58	51.60	52.89	45.61	45.95	44.53	44.55	52.38
musiq-koniq⁵⁵	62.28	61.37	62.82	60.04	56.91	57.19	56.41	56.43	68.46
MANIQA⁹⁴	0.66	0.64	0.66	0.61	0.62	0.62	0.61	0.63	0.63
CLIPIQA⁹⁵	0.56	0.52	0.55	0.49	0.56	0.55	0.54	0.56	0.56
TReS-koniq⁹⁶	69.64	64.48	70.45	70.33	65.16	65.19	64.22	64.24	66.42
HyperIQA⁹⁷	0.53	0.52	0.54	0.53	0.44	0.44	0.42	0.44	0.45
GPR-BIQA¹⁰	0.70	0.68	0.69	0.67	0.61	0.61	0.60	0.62	0.72
Quality Net⁹⁸	0.72	0.70	0.65	0.69	0.63	0.63	0.62	0.63	0.74
PIQI⁹	0.71	0.69	0.64	0.68	0.62	0.62	0.61	0.60	0.73
Average	24.21	23.71	24.43	24.18	22.62	22.67	22.43	22.44	24.76

Quick links

Search