CleanNet: Transfer Learning for Scalable Image Classifier Training with Label Noise.
Kuang-Huei LeeXiaodong HeLei ZhangLinjun YangPublished in: CoRR (2017)
Keyphrases
- transfer learning
- classifier training
- image classification
- input image
- learning tasks
- image segmentation
- multi class
- image retrieval
- image features
- semi supervised learning
- image content
- multi task learning
- active learning
- multiscale
- dictionary learning
- labeled data
- cross domain
- image regions
- image collections
- similarity measure
- feature selection
- training data
- text categorization
- text classification
- training set
- probabilistic model