Predicting Class Distribution Shift for Reliable Domain Adaptive Object Detection.
Nicolas Harvey ChapmanFeras DayoubWill BrowneChristopher F. LehnertPublished in: CoRR (2023)
Keyphrases
- class distribution
- object detection
- training data
- class imbalance
- training set
- test set
- highly imbalanced
- highly skewed
- imbalanced data sets
- imbalanced datasets
- cost sensitive
- training examples
- computer vision
- class labels
- multi class
- training set size
- object recognition
- unlabeled data
- nearest neighbor
- cost sensitive learning
- pairwise
- decision trees