Positive-Unlabeled Data Purification in the Wild for Object Detection.
Jianyuan GuoKai HanHan WuChao ZhangXinghao ChenChunjing XuChang XuYunhe WangPublished in: CVPR (2021)
Keyphrases
- unlabeled data
- object detection
- labeled data
- semi supervised learning
- semi supervised
- co training
- active learning
- training data
- supervised learning
- text classification
- labeled and unlabeled data
- semi supervised classification
- learning algorithm
- labeled training data
- labeled examples
- computer vision
- positive and negative
- training examples
- number of labeled examples
- text categorization
- training set
- class labels
- object class
- object categories
- domain adaptation
- data points
- supervised learning algorithms
- multi class
- small set of labeled
- object recognition
- knowledge discovery
- single view
- multi view
- prior knowledge
- label propagation
- transfer learning
- multi view learning
- decision trees