Cost-Sensitive Uncertainty-Based Failure Recognition for Object Detection.
Moussa Kassem SbeytiMichelle KargChristian WirthNadja KleinSahin AlbayrakPublished in: CoRR (2024)
Keyphrases
- cost sensitive
- object detection
- multi class
- misclassification costs
- cost sensitive classification
- object recognition
- cost sensitive learning
- binary classification
- naive bayes
- face detection
- class distribution
- boosting algorithms
- support vector machine
- object class
- class imbalance
- active learning
- pairwise
- object categories
- probability estimates
- feature extraction
- small number
- training data