Labels Are Not Perfect: Inferring Spatial Uncertainty in Object Detection.
Di FengZining WangYiyang ZhouLars RosenbaumFabian TimmKlaus DietmayerMasayoshi TomizukaWei ZhanPublished in: CoRR (2020)
Keyphrases
- object detection
- training data
- spatial information
- spatial data
- object class
- multi class
- space time
- object recognition
- face detection
- spatial distribution
- class labels
- computer vision
- scene understanding
- object classification
- learning algorithm
- pedestrian detection
- spatial and temporal
- training set
- data sets
- multi label
- detection method
- spatial relationships
- uncertain data
- training examples
- pairwise
- support vector
- spatial reasoning
- human detection
- robust optimization
- image labeling