Leveling Down in Computer Vision: Pareto Inefficiencies in Fair Deep Classifiers.
Dominik ZietlowMichael LohausGuha BalakrishnanMatthäus KleindessnerFrancesco LocatelloBernhard SchölkopfChris RussellPublished in: CoRR (2022)
Keyphrases
- computer vision
- multi objective
- training data
- object recognition
- multi objective optimization
- image processing
- data sets
- machine learning algorithms
- image understanding
- object detection
- classification systems
- pattern recognition
- classification algorithm
- linear classifiers
- computer graphics
- image segmentation
- classification method
- computer vision algorithms
- supervised classification
- image matching
- naive bayes
- feature set
- optical flow
- machine learning
- machine learning methods
- shape matching
- extracted features
- class labels
- test set
- image sequences
- support vector
- image analysis
- training set
- multicriteria optimization
- multiple classifiers
- individual classifiers
- feature selection
- ensemble classifier
- ensemble learning
- evolutionary algorithm
- vision system
- classification models
- training examples
- action recognition
- structure from motion