Active learning for deep object detection by fully exploiting unlabeled data.
Feixiang TanGuansheng ZhengPublished in: Connect. Sci. (2023)
Keyphrases
- unlabeled data
- active learning
- object detection
- labeled data
- semi supervised
- semi supervised learning
- unlabeled examples
- supervised learning
- co training
- training set
- training examples
- labeled examples
- semi supervised classification
- learning algorithm
- random sampling
- labeled and unlabeled data
- labeled training data
- class labels
- transfer learning
- selective sampling
- number of labeled examples
- machine learning
- computer vision
- object categories
- text categorization
- unlabeled instances
- object class
- multi class
- data sets
- text classification
- class distribution
- labeling effort
- positive examples
- domain adaptation
- training data
- labeled data for training
- sample selection
- supervised learning algorithms
- object recognition
- learning process
- prior knowledge
- data points
- decision trees
- labeled instances
- pairwise
- active learner