1.4 《Learning with Bounded Instance- and label-dependent Label Noise》 This paper focus on Bounded Instance- and Label-dependent label Noise (BILN), a particular case of Label-dependent label Noise where the label noise rates. This paper focus on a particular case of ILN where noise rates h...
CVPR2019: Probabilistic End-to-end Noise Correction for Learning with Noisy Labels 对上篇文章的改进 用模型去学习样本真实标签的分布,Joint Training,不再做两步更新 1.2 Dataset Pruning 直接移除噪声数据,同样可以达到“清洗数据,使用干净数据训练分类器”的目的 ICCV2019: O2U-Net: A Simple Noisy Label Dete...
Existing research on learning with noisy labels predominantly focuses on synthetic label noise. Paper Add Code Foster Adaptivity and Balance in Learning with Noisy Labels no code yet • 3 Jul 2024 Moreover, existing methods tend to neglect the class balance in selecting samples, leading to bias...
基于概率模型(estimate noisy label) 包括EM-based 模型、置信学习等。 基本的数学模型是: noise 与 label 有关,狮子容易被分类成猫,但不容易被分类为轮船。 找noisy label 和 true label 之间联合概率分布矩阵、转移矩阵。 用概率矩阵...
A curated list of resources for Learning with Noisy Labels - JUNXYU/Awesome-Learning-with-Label-Noise
A curated list of resources for Learning with Noisy Labels - congyang1996/Awesome-Learning-with-Label-Noise
Learning with symmetric label noise: The importance of being unhinged. In NIPS*29, 2015.Brendan Van Rooyen, Aditya Menon, and Robert C Williamson. Learning with symmetric label noise: The importance of being unhinged. In NIPS, pages 10-18, 2015....
However, we observe that the advantage of LS vanishes when we operate in a high label noise regime. Puzzled by the observation, we proceeded to discover that several proposed learning-with-noisy-labels solutions in the literature instead relate more closely to negative label smoothing (NLS), ...
However, Long and Servedio [2010] proved that under symmetric label noise (SLN), minimisation of any convex potential over a linear function class can result in classification performance equivalent to random guessing. This ostensibly shows that convex losses are not SLN-robust. In this paper, we...
Recently, noise-cleansing methods have been considered as the most competitive noisy-label learning algorithms. Despite their success, their noisy label detectors are often based on heuristics more than a theory, requiring a robust classifier to predict the noisy data with loss values. In this ...