Two-phase training mitigates class imbalance for camera trap image classification with CNNs.
Farjad MalikSimon WoutersRuben CartuyvelsErfan GhaderyMarie-Francine MoensPublished in: CoRR (2021)
Keyphrases
- class imbalance
- image classification
- class distribution
- training set
- active learning
- class imbalanced
- cost sensitive
- test set
- sampling methods
- cost sensitive learning
- imbalanced datasets
- high dimensionality
- minority class
- feature extraction
- majority class
- imbalanced data
- concept drift
- machine learning
- training examples
- support vector
- small disjuncts
- feature selection
- training dataset
- decision boundary
- classification error
- supervised learning
- high dimensional