Sampling Based Average Classifier Fusion

Table 2

Flower-17 and Caltech-101 recognition rates and comparison.

Flower-17		Caltech-101
Method	Accuracy	Method	Accuracy
Best single	75.1±1.5	Best single	66.2±1.2
Average1	77.5±1.7	Average1	58.8±1.7
Average2	86.0±1.5	Average2	71.0±1.2
[12]	𝟖𝟖.𝟑±𝟎.𝟑	[13]	𝟕𝟕.𝟖±𝟎.𝟒
[13]	85.5±3.0	[14]	66.2±0.5