AdaCos: Adaptively Scaling Cosine Logits for Effectively Learning Deep Face Representations

CVPR 2019 Xiao ZhangRui ZhaoYu QiaoXiaogang WangHongsheng Li

The cosine-based softmax losses and their variants achieve great success in deep learning based face recognition. However, hyperparameter settings in these losses have significant influences on the optimization path as well as the final recognition performance... (read more)

PDF Abstract

Results from the Paper


TASK DATASET MODEL METRIC NAME METRIC VALUE GLOBAL RANK RESULT BENCHMARK
Face Verification Labeled Faces in the Wild Dyna. AdaCos Accuracy 99.73% # 5
Face Verification MegaFace Dynamic AdaCos Accuracy 97.41% # 2

Methods used in the Paper