A Computational Approach towards Visual Object Recognition at Taxonomic Levels of Concepts

Table 4

Comparison results of classification on basic conceptual categories. Results are averaged over 10 iterations. Time complexity is averaged over all train samples.

Features	Dataset 1				Dataset 3				#Feature vector dimensions
	Accuracy			Average processing time per sample	Accuracy			Average processing time per sample
	Animal class	Plant class	Total (over all test samples)		Animal class	Plant class	Total (over all test samples)
C2	86.55; (2.66)	84.33; (2.98)	85.27; (2.41)	3.56	93.50; (2.28)	95.50; (3.14)	94.50; (1.48)	7.23	200
HOG	82.37; (3.21)	80.33; (1.62)	81.13; (1.57)	0.0385	85.66; (4.02)	94.66; (2.58)	90.16; (2.03)	0.1452	128
Moment-based method	86.67; (2.57)	85.35; (2.78)	85.71; (.79)	0.1465	94.66; (3.16)	95.16; (1.99)	94.92; (1.93)	0.2902	32