How to Overcome Confirmation Bias in Semi-Supervised Image Classification by Active Learning.
Sandra GilhuberRasmus HvingelbyMang Ling Ada FokThomas SeidlPublished in: ECML/PKDD (2) (2023)
Keyphrases
- semi supervised
- active learning
- image classification
- labeled data
- unlabeled data
- semi supervised learning
- supervised learning
- image representation
- co training
- unsupervised learning
- metric learning
- learning algorithm
- training set
- bag of words
- pool based active learning
- image features
- semi supervised clustering
- pairwise
- labeled examples
- experimental design
- random sampling
- transfer learning
- semi supervised classification
- sparse representation
- learning strategies
- training examples
- multi view
- learning process
- machine learning
- sparse coding
- data sets
- batch mode
- class specific
- subspace learning
- label propagation
- relevance feedback
- sample selection
- unlabeled instances
- fully supervised