Decoupling Classifier for Boosting Few-shot Object Detection and Instance Segmentation.
Bin-Bin GaoXiaochen ChenZhongyi HuangCongchong NieJun LiuJinxiang LaiGuannan JiangXi WangChengjie WangPublished in: NeurIPS (2022)
Keyphrases
- object detection
- discriminative classifiers
- boosted classifiers
- face detection
- object segmentation
- image parsing
- adaboost algorithm
- weak classifiers
- weak learners
- feature selection
- ensemble learning
- learning algorithm
- pixel classification
- segmentation method
- background subtraction
- object class
- face detector
- segmentation algorithm
- image segmentation
- object detectors
- object categories
- level set
- multiscale
- computer vision
- temporal segmentation
- scene understanding
- video sequences
- object recognition
- training data
- bounding box
- video segmentation
- feature space
- video content
- support vector machine
- ensemble classifier
- multiclass classification
- medical images
- decision stumps
- multiple classifier systems
- improving classification accuracy
- boosting algorithms
- video shots
- base classifiers
- input output
- classification algorithm
- class labels
- feature set
- decision trees