Support Vector Guided Softmax Loss for Face Recognition

29 Dec 2018Xiaobo WangShuo WangShifeng ZhangTianyu FuHailin ShiTao Mei

Face recognition has witnessed significant progresses due to the advances of deep convolutional neural networks (CNNs), the central challenge of which, is feature discrimination. To address it, one group tries to exploit mining-based strategies (\textit{e.g.}, hard example mining and focal loss) to focus on the informative examples... (read more)

PDF Abstract
TASK DATASET MODEL METRIC NAME METRIC VALUE GLOBAL RANK RESULT BENCHMARK
Face Identification MegaFace SV-AM-Softmax Accuracy 97.2% # 2
Face Verification MegaFace SV-AM-Softmax Accuracy 97.38% # 3
Face Verification Trillion Pairs Dataset SV-AM-Softmax Accuracy 72.71 # 1
Face Identification Trillion Pairs Dataset SV-AM-Softmax Accuracy 73.56 # 1

Methods used in the Paper