Noise-Tolerant Paradigm for Training Face Recognition CNNs

03/25/2019
by   Yangyu Huang, et al.
0

Benefit from large-scale training datasets, deep Convolutional Neural Networks(CNNs) have achieved impressive results in face recognition(FR). However, tremendous scale of datasets inevitably lead to noisy data, which obviously reduce the performance of the trained CNN models. Kicking out wrong labels from large-scale FR datasets is still very expensive, even some cleaning approaches are proposed. According to the analysis of the whole process of training CNN models supervised by angular margin based loss(AM-Loss) functions, we find that the θ distribution of training samples implicitly reflects their probability of being clean. Thus, we propose a novel training paradigm that employs the idea of weighting samples based on the above probability. Without any prior knowledge of noise, we can train high performance CNN models with large-scale FR datasets. Experiments demonstrate the effectiveness of our training paradigm. The codes are available at https://github.com/huangyangyu/NoiseFace.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
01/23/2018

ArcFace: Additive Angular Margin Loss for Deep Face Recognition

Convolutional neural networks have significantly boosted the performance...
research
04/26/2017

Deep Convolutional Neural Network to Detect J-UNIWARD

This paper presents an empirical study on applying convolutional neural ...
research
08/16/2021

Pixel Difference Networks for Efficient Edge Detection

Recently, deep Convolutional Neural Networks (CNNs) can achieve human-le...
research
04/15/2022

Deep Unlearning via Randomized Conditionally Independent Hessians

Recent legislation has led to interest in machine unlearning, i.e., remo...
research
09/29/2021

Improvising the Learning of Neural Networks on Hyperspherical Manifold

The impact of convolution neural networks (CNNs) in the supervised setti...
research
11/09/2015

A Light CNN for Deep Face Representation with Noisy Labels

Convolution neural network (CNN) has significantly pushed forward the de...
research
09/30/2021

Learning to Predict Trustworthiness with Steep Slope Loss

Understanding the trustworthiness of a prediction yielded by a classifie...

Please sign up or login with your details

Forgot password? Click here to reset