. Author manuscript; available in PMC: 2014 Nov 21.

Published in final edited form as: Proc IEEE Int Conf Comput Vis. 2013 Dec;2013:2168–2175. doi: 10.1109/ICCV.2013.269

Table 3.

Testing performance of different methods for the Weiz- mann horse dataset.

Method	F-value	G-mean	Pixel accuracy
KSSVM [3]	-	-	94.60%
TWM [17]	-	-	94.70%
Auto-context [28]	84%	-	-
Levin & Weiss [21]	-	-	95.2%
MSANN [24]	87.58%	92.76%	94.34%
CHM-RF	83.15%	90.20%	92.33%
CHM-LDNN	89.89%	94.39%	95.37%