Loss Re-Scaling VQA: Revisiting the Language Prior Problem From a Class-Imbalance View.
Yangyang GuoLiqiang NieZhiyong ChengQi TianMin ZhangPublished in: IEEE Trans. Image Process. (2022)
Keyphrases
- class imbalance
- class distribution
- active learning
- cost sensitive
- cost sensitive learning
- high dimensionality
- small disjuncts
- majority class
- sampling methods
- feature selection
- imbalanced datasets
- minority class
- concept drift
- software defect prediction
- imbalanced class distribution
- prior knowledge
- machine learning
- change detection
- semi supervised
- training set
- learning algorithm