Utilizing dataset affinity prediction in object detection to assess training data.
Stefan BeckerJens BayerRonny HugWolfgang HübnerMichael ArensPublished in: CoRR (2023)
Keyphrases
- object detection
- training data
- pascal voc
- training dataset
- prediction accuracy
- test data
- object detectors
- prediction model
- object categories
- decision trees
- manually labeled training data
- data sets
- face detection
- domain knowledge
- object recognition
- prior knowledge
- training examples
- object classification
- benchmark datasets
- training set
- weakly labeled
- learning algorithm
- classification accuracy
- class labels
- prediction error
- prediction algorithm
- affinity measure
- object segmentation
- human actions
- test set
- input data
- computer vision
- object class
- pedestrian detection
- background subtraction
- representative set
- multi class
- neural network