Skip to main content
. 2020 Nov 6;4:168. [Version 1] doi: 10.12688/gatesopenres.13197.1

Table 4. Top-1 and Top-10 identification accuracies of the MOS dataset (Cohort 2, N=238) at different resizing settings using our experimental platform (Python).

Width
(pixels)		 Top-1
Accuracy		 Top-10
Accuracy
75 77.73% 88.65%
100 77.73% 92.01%
125 77.73% 88.65%
150 80.25% 89.07%
175 73.94% 89.07%
200 76.47% 87.81%