Boosting Soft-Margin SVM with Feature Selection for Pedestrian Detection.
Kenji NishidaTakio KuritaPublished in: Multiple Classifier Systems (2005)
Keyphrases
- soft margin
- pedestrian detection
- support vector
- feature selection
- support vector machine
- human body
- object detection
- generalization ability
- multi class
- boosting algorithms
- svm classifier
- support vector machine svm
- support vectors
- detection rate
- parameter selection
- text classification
- hyperplane
- model selection
- binary classification
- benchmark datasets
- svm classification
- text categorization
- kernel function
- cross validation
- feature extraction
- classification accuracy
- machine learning
- perceptron algorithm
- training data
- generalization error
- histogram intersection kernel
- loss function
- knn
- feature vectors
- multi label
- face detection
- training examples
- detection algorithm
- feature set
- training set
- three dimensional
- linearly separable
- computer vision
- learning algorithm