Towards Better Uncertainty Sampling: Active Learning with Multiple Views for Deep Convolutional Neural Network.
Tao HeXiaoming JinGuiguang DingLan YiChenggang YanPublished in: ICME (2019)
Keyphrases
- multiple views
- uncertainty sampling
- convolutional neural network
- active learning
- cost sensitive
- multi view
- experimental design
- face detection
- random sampling
- viewpoint
- single view
- d objects
- semi supervised
- query by committee
- training examples
- neural network
- supervised learning
- learning algorithm
- training set
- misclassification costs
- machine learning
- range images
- learning strategies
- labeled data
- ensemble members
- maximum entropy
- ensemble methods
- unlabeled data
- generalization error
- empirical studies
- learning process
- logistic regression
- detection method
- semi supervised learning
- minimize total
- learning environment
- three dimensional