paper-with-me

홈 › Papers

Understanding and Utilizing Deep Neural Networks Trained with Noisy Labels

2019-05-13 · Pengfei Chen, Benben Liao, Guangyong Chen, Shengyu Zhang

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 noisy labels. In this paper, we find that the test accuracy can be quantitatively characterized in terms of the noise ratio in datasets. In particular, the test accuracy is a quadratic function of the noise ratio in the case of symmetric noise, which explains the experimental findings previously published. Based on our analysis, we apply cross-validation to randomly split noisy datasets, which identifies most samples that have correct labels. Then we adopt the Co-teaching strategy which takes full advantage of the identified samples to train DNNs robustly against noisy labels. Compared with extensive state-of-the-art methods, our strategy consistently improves the generalization performance of DNNs under both synthetic and real-world training noise.

📄 PDF Abstract BibTeX arXiv:1905.05040

Code (3)

chenpf1025/noisy_label_understanding_utilizing 공식 구현 tf
machine-learning-class-project/machine-learning-class-project.github.io
noise-learning/selfmix pytorch

Tasks

Image Classification

Similar Papers 제목 키워드 기반

Self-refining of Pseudo Labels for Music Source Separation with Noisy Labeled Data

2023-07-24 · Junghyun Koo, Yunkee Chae, Chang-Bin Jeon, Kyogu Lee

Music source separation (MSS) faces challenges due to the limited availability of correctly-labeled individual instrument tracks. With the push to acquire larger datasets to improve MSS performance, the inevitability of …

Instrument RecognitionMusic Source Separation

Knowledge Distillation with Noisy Labels for Natural Language Understanding

2021-09-21 · WNUT (ACL) 2021 11 · Shivendra Bhardwaj, Abbas Ghaddar, Ahmad Rashid, Khalil Bibi 외

Knowledge Distillation (KD) is extensively used to compress and deploy large pre-trained language models on edge devices for real-world applications. However, one neglected area of research is the impact of noisy (corrup…

Knowledge DistillationNatural Language Understanding

On the Robustness of Monte Carlo Dropout Trained with Noisy Labels

2021-03-22 · Purvi Goel, Li Chen

The memorization effect of deep learning hinders its performance to effectively generalize on test set when learning with noisy labels. Prior study has discovered that epistemic uncertainty techniques are robust when tra…

Learning with noisy labelsMemorization

Dimensionality-Driven Learning with Noisy Labels

2018-06-07 · ICML 2018 7 · Xingjun Ma, Yisen Wang, Michael E. Houle, Shuo Zhou 외

Datasets with significant proportions of noisy (incorrect) class labels present challenges for training accurate Deep Neural Networks (DNNs). We propose a new perspective for understanding DNN generalization for such dat…

Image ClassificationLearning with noisy labels

Deep Neural Networks Learn Meta-Structures from Noisy Labels in Semantic Segmentation

2021-03-22 · Yaoru Luo, Guole Liu, Yuanhao Guo, Ge Yang

How deep neural networks (DNNs) learn from noisy labels has been studied extensively in image classification but much less in image segmentation. So far, our understanding of the learning behavior of DNNs trained by nois…

image-classificationImage ClassificationImage SegmentationSegmentation+1