แนวคิดหลัก
Utilizing 3D spatio-spatial temporal convolution, FH-SSTNet offers superior performance in forehead-based user verification.
สถิติ
Experimentation on the FH-V1 dataset shows superior performance of FH-SSTNet compared to ResNet50.
Proposed method achieved an EER of 1.77% with TMR and FMR at 1.15% and 0.15%.
คำพูด
"Our proposed algorithm, FH-SSTNet, achieved an EER of 32.73% with TMR and FMR at 1.15% and 0.15%, respectively."
"The improved performance of the proposed method can be attributed to the proposed 3D representation in FH-SSTNet."