BriefGPT.xyz
May, 2019
利用带噪标签训练的深度神经网络及其应用
Understanding and Utilizing Deep Neural Networks Trained with Noisy Labels
HTML
PDF
Pengfei Chen, Benben Liao, Guangyong Chen, Shengyu Zhang
TL;DR
本文针对网络训练时噪声标签的问题，提出了一种基于交叉验证和合作训练策略的解决方法，成功提高了深度神经网络在合成和实际噪音数据集下的泛化性能。
Abstract
noisy labels
are ubiquitous in real-world datasets, which poses a challenge for robustly training
deep neural networks
(DNNs) as DNNs usually have the high capacity to memorize the
→