Rethinking Pseudo Labels for Semi-Supervised Object Detection.
Hengduo LiZuxuan WuAbhinav ShrivastavaLarry S. DavisPublished in: CoRR (2021)
Keyphrases
- semi supervised
- object detection
- label propagation
- pairwise
- fully supervised
- label information
- unlabeled samples
- semi supervised learning
- labeled examples
- pairwise constraints
- partially labeled
- labeled data
- unlabeled data
- partially labeled data
- labeled samples
- weakly supervised
- object class
- multi label classification
- unsupervised learning
- computer vision
- supervised learning
- pascal voc
- active learning
- fully labeled
- pedestrian detection
- face detection
- object categories
- unlabeled instances
- multi class
- multi label learning
- metric learning
- object classification
- scene understanding
- object recognition
- class labels
- multi view
- scene recognition
- semi supervised classification
- instance level
- semantic segmentation
- object segmentation
- multi label
- constrained clustering
- generative model
- training examples
- background subtraction
- training data
- machine learning
- training samples
- multiple instance learning