Boosting-based Face Detection and AdaptationFace detection, because of its vast array of applications, is one of the most active research areas in computer vision. In this book, we review various approaches to face detection developed in the past decade, with more emphasis on boosting-based learning algorithms. We then present a series of algorithms that are empowered by the statistical view of boosting and the concept of multiple instance learning. We start by describing a boosting learning framework that is capable to handle billions of training examples. It differs from traditional bootstrapping schemes in that no intermediate thresholds need to be set during training, yet the total number of negative examples used for feature selection remains constant and focused (on the poor performing ones). A multiple instance pruning scheme is then adopted to set the intermediate thresholds after boosting learning. This algorithm generates detectors that are both fast and accurate. Table of Contents: A Brief Survey of the Face Detection Literature / Cascade-based Real-Time Face Detection / Multiple Instance Learning for Face Detection / Detector Adaptation / Other Applications / Conclusions and Future Work |
Contents
1 | |
Cascadebased RealTime Face Detection | 29 |
Multiple Instance Learning for Face Detection | 45 |
Detector Adaptation | 69 |
Other Applications | 83 |
Conclusions and Future Work | 111 |
Bibliography | 113 |
Authors Biographies | 127 |
Other editions - View all
Common terms and phrases
AdaBoost AnyBoost approach BMSD boosting algorithm boosting classifier boosting learning boosting-based celebrity classifier adaptation clustering computed Computer Vision cost function CVPR detection rate detection window detector face detection face recognition false positive rate feature selection Figure filters final threshold ground truth Haar features Haar-like features histograms IEEE Trans improve labeled frames learning algorithm likelihood logistic regression machine learning manual labels MILBoost multi-view face multiple instance learning multiview face detection negative examples neural network node number of weak object detection optimal output overfitting parameters performance person detection pixels pose positive examples positive windows Proc proposed rectangle rejection thresholds ROC curves sampling scheme score shown in Fig similarity labels soft cascade speaker detection SSL+MPD subcategory classifiers support vector machines total number training examples training set Update vector Viola and Jones weak classifiers weight trimming WTA-McBoost Z score Zhang