As an emerging topic in face recognition, designing margin-based loss
functions can increase the feature margin between different classes for
enhanced discriminability. More recently, the idea of mining-based strategies
is adopted to emphasize the misclassified samples, achieving promising results.
However, during the entire training process, the prior methods either do not
explicitly emphasize the sample based on its importance that renders the hard
samples not fully exploited; or explicitly emphasize the effects of
semi-hard/hard samples even at the early training stage that may lead to
convergence issue. In this work, we propose a novel Adaptive Curriculum
Learning loss (CurricularFace) that embeds the idea of curriculum learning into
the loss function to achieve a novel training strategy for deep face
recognition, which mainly addresses easy samples in the early training stage
and hard ones in the later stage. Specifically, our CurricularFace adaptively
adjusts the relative importance of easy and hard samples during different
training stages. In each stage, different samples are assigned with different
importance according to their corresponding difficultness. Extensive
experimental results on popular benchmarks demonstrate the superiority of our
CurricularFace over the state-of-the-art competitors.
Описание
CurricularFace: Adaptive Curriculum Learning Loss for Deep Face Recognition
%0 Generic
%1 huang2020curricularface
%A Huang, Yuge
%A Wang, Yuhan
%A Tai, Ying
%A Liu, Xiaoming
%A Shen, Pengcheng
%A Li, Shaoxin
%A Li, Jilin
%A Huang, Feiyue
%D 2020
%K computer-vision face-recognition
%T CurricularFace: Adaptive Curriculum Learning Loss for Deep Face
Recognition
%U http://arxiv.org/abs/2004.00288
%X As an emerging topic in face recognition, designing margin-based loss
functions can increase the feature margin between different classes for
enhanced discriminability. More recently, the idea of mining-based strategies
is adopted to emphasize the misclassified samples, achieving promising results.
However, during the entire training process, the prior methods either do not
explicitly emphasize the sample based on its importance that renders the hard
samples not fully exploited; or explicitly emphasize the effects of
semi-hard/hard samples even at the early training stage that may lead to
convergence issue. In this work, we propose a novel Adaptive Curriculum
Learning loss (CurricularFace) that embeds the idea of curriculum learning into
the loss function to achieve a novel training strategy for deep face
recognition, which mainly addresses easy samples in the early training stage
and hard ones in the later stage. Specifically, our CurricularFace adaptively
adjusts the relative importance of easy and hard samples during different
training stages. In each stage, different samples are assigned with different
importance according to their corresponding difficultness. Extensive
experimental results on popular benchmarks demonstrate the superiority of our
CurricularFace over the state-of-the-art competitors.
@misc{huang2020curricularface,
abstract = {As an emerging topic in face recognition, designing margin-based loss
functions can increase the feature margin between different classes for
enhanced discriminability. More recently, the idea of mining-based strategies
is adopted to emphasize the misclassified samples, achieving promising results.
However, during the entire training process, the prior methods either do not
explicitly emphasize the sample based on its importance that renders the hard
samples not fully exploited; or explicitly emphasize the effects of
semi-hard/hard samples even at the early training stage that may lead to
convergence issue. In this work, we propose a novel Adaptive Curriculum
Learning loss (CurricularFace) that embeds the idea of curriculum learning into
the loss function to achieve a novel training strategy for deep face
recognition, which mainly addresses easy samples in the early training stage
and hard ones in the later stage. Specifically, our CurricularFace adaptively
adjusts the relative importance of easy and hard samples during different
training stages. In each stage, different samples are assigned with different
importance according to their corresponding difficultness. Extensive
experimental results on popular benchmarks demonstrate the superiority of our
CurricularFace over the state-of-the-art competitors.},
added-at = {2022-05-28T09:45:18.000+0200},
author = {Huang, Yuge and Wang, Yuhan and Tai, Ying and Liu, Xiaoming and Shen, Pengcheng and Li, Shaoxin and Li, Jilin and Huang, Feiyue},
biburl = {https://www.bibsonomy.org/bibtex/215df74f978fa60940b97e92ce25cf41c/adnanahmed},
description = {CurricularFace: Adaptive Curriculum Learning Loss for Deep Face Recognition},
interhash = {edb7b54887789dadfd9053868cccfd16},
intrahash = {15df74f978fa60940b97e92ce25cf41c},
keywords = {computer-vision face-recognition},
note = {cite arxiv:2004.00288Comment: CVPR 2020},
timestamp = {2022-05-31T23:05:01.000+0200},
title = {CurricularFace: Adaptive Curriculum Learning Loss for Deep Face
Recognition},
url = {http://arxiv.org/abs/2004.00288},
year = 2020
}