Boosting with cross-validation based feature selection for pedestrian detection.
Kenji NishidaTakio KuritaPublished in: IJCNN (2008)
Keyphrases
- cross validation
- pedestrian detection
- feature selection
- model selection
- support vector
- human body
- selected features
- object detection
- variable selection
- detection rate
- hyperparameters
- human detection
- regression problems
- classification accuracy
- text classification
- cross validated
- machine learning
- histograms of oriented gradients
- feature set
- generalization error
- base learners
- ensemble learning
- multi class
- training set
- leave one out cross validation
- benchmark datasets
- feature space
- error estimates
- naive bayes
- feature subset
- pose estimation
- multi task
- data mining
- support vector machine
- knn
- meta learning
- computer vision
- three dimensional
- feature extraction
- viewpoint
- detection algorithm
- face detection
- base classifiers