Research Article

Eye Movement Signal Classification for Developing Human-Computer Interface Using Electrooculogram

Table 2

Performance accuracy of convolution features using DTDNN.

S. no.Sub.Hidden neuronMean training time (sec)Mean testing time (sec)Recognizing accuracy
MaxMinMeanStd.

1S183.510.6793.6486.3690.412.14
2S283.560.6193.7285.5690.491.88
3S383.570.6393.7486.3690.402.25
4S483.700.6592.7486.3690.581.79
5S583.480.6293.6485.4590.322.83
6S683.590.7693.6485.5690.131.99
7S783.620.6293.6485.4590.592.63
8S883.620.6293.6486.3690.491.64
9S983.720.6593.6485.4590.342.00
10S1083.590.6393.6486.3690.232.31
11S1183.440.6194.5585.5690.472.23
12S1283.550.6394.5588.1891.761.79
13S1383.710.6394.5586.3690.632.13
14S1483.920.6194.5585.4590.862.26
15S1583.680.6393.6486.3690.501.97
16S1683.600.6394.5585.5590.592.08
17S1783.820.7393.6486.3690.541.87
18S1883.650.7093.6486.3690.592.17
19S1983.670.7393.6485.4590.581.79
20S2083.840.7394.5586.3690.732.50