BoVW model based on adaptive local and global visual words modeling and log-based relevance feedback for semantic retrieval of the images

EURASIP Journal on Image and Video Processing

Table 3 Comparative analysis of competitive methods with a proposed method on the Corel 1000 dataset

Semantic category		Color-shape-texture [24]	CNN-RCMF-ICM [21]	DTSD-MR [22]	Region-texture [23]	LBPC-LBPH-CH [2]	MTBCD [13]	Proposed method
African tribes	P	85.00	66.30	64.33	81.00	61.52	66.00	81.36
	R	17.00	13.26	06.43	16.00	12.30	13.20	16.27
	F	28.33	22.10	11.69	26.72	20.50	22.00	27.12
Beaches	P	75.00	52.70	28.82	75.00	43.36	64.00	78.74
	R	15.00	10.54	02.88	15.00	08.67	12.80	15.74
	F	25.00	17.57	05.24	25.00	14.45	21.33	26.24
Building	P	60.00	70.60	54.23	67.00	54.57	82.00	82.93
	R	12.00	14.12	05.42	13.00	10.91	16.40	16.58
	F	20.00	23.53	09.86	21.78	18.18	27.33	27.64
Buses	P	100.0	99.10	79.43	93.00	87.90	100.0	89.56
	R	20.00	19.82	07.94	19.00	17.58	20.00	17.91
	F	33.33	33.03	14.44	31.55	29.30	33.33	29.85
Dinosaurs	P	100.0	100.0	99.59	97.00	98.88	100.0	100.0
	R	20.00	20.00	09.96	19.00	19.78	20.00	20.00
	F	33.33	33.33	18.11	31.78	32.97	33.33	33.33
Elephants	P	80.00	90.80	40.09	91.00	44.54	83.00	93.09
	R	16.00	18.16	04.01	18.00	08.91	16.60	18.61
	F	26.67	30.27	07.29	30.06	14.85	27.67	31.03
Flowers	P	100.0	99.20	78.25	90.00	83.08	94.00	97.11
	R	20.00	19.84	07.83	18.00	16.62	18.80	19.42
	F	33.33	33.07	14.24	30.00	27.70	31.33	32.37
Horses	P	90.00	95.10	71.17	84.00	81.59	99.00	96.35
	R	18.00	19.02	07.12	17.00	16.32	19.80	19.27
	F	30.00	31.70	12.94	28.28	27.20	33.00	32.11
Mountain	P	85.00	71.50	37.55	59.00	39.89	67.00	86.73
	R	17.00	14.30	03.76	12.00	07.98	13.40	17.34
	F	28.33	23.83	06.84	19.94	13.30	22.33	28.91
Foods	P	85.00	85.30	58.40	79.00	56.31	95.00	93.24
	R	17.00	17.06	05.84	16.00	11.26	19.00	18.64
	F	28.33	28.43	10.62	26.61	18.77	31.67	31.08
mAP (%) Avg. R Avg. F	P	86.00	83.06	61.19	81.60	65.16	85.00	89.91
	R	17.20	16.61	06.12	16.30	13.03	17.00	17.98
	F	28.66	27.68	11.12	27.17	21.72	28.33	29.97