PNP: Robust Learning from Noisy Labels by Probabilistic Noise Prediction.
Zeren SunFumin ShenDan HuangQiong WangXiangbo ShuYazhou YaoJinhui TangPublished in: CVPR (2022)
Keyphrases
- label noise
- noisy data
- learning algorithm
- learning process
- training data
- prediction accuracy
- neural network
- noise tolerant
- noisy environments
- impulse noise
- background knowledge
- data sets
- artificial neural networks
- prior knowledge
- learning tasks
- machine learning
- geometric distortions
- e learning
- decision trees
- gaussian noise
- median filter
- training set
- learning problems
- multi label
- pairwise
- missing data
- generative model
- knowledge acquisition
- hidden markov models
- active learning