1 code implementation • 13 Oct 2022 • Haiyu Wu, Grace Bezold, Manuel Günther, Terrance Boult, Michael C. King, Kevin W. Bowyer
Two annotators independently assigning attribute values shows that only 12 of 40 common attributes are assigned values with >= 95% consistency, and three (high cheekbones, pointed nose, oval face) have essentially random consistency.
no code implementations • 11 Nov 2018 • Ahmed ElSayed, Elif Kongar, Ausif Mahmood, Tarek Sobh, Terrance Boult
Using heterogeneous depth cameras and 3D scanners in 3D face verification causes variations in the resolution of the 3D point clouds.
3 code implementations • 22 Mar 2016 • Ethan Rudd, Manuel Günther, Terrance Boult
Unfortunately, it can be difficult to apply joint optimization to DCNNs when training data is imbalanced, and re-balancing multi-label data directly is structurally infeasible, since adding/removing data to balance one label will change the sampling of the other labels.
4 code implementations • CVPR 2016 • Abhijit Bendale, Terrance Boult
We present a methodology to adapt deep networks for open set recognition, by introducing a new model layer, OpenMax, which estimates the probability of an input being from an unknown class.
no code implementations • CVPR 2015 • Abhijit Bendale, Terrance Boult
We prove that thresholding sums of monotonically decreasing functions of distances in linearly transformed feature space can balance "open space risk" and empirical risk.