计算机应用研究Issue(11)：3517-3520,4.DOI:10.3969/j.issn.1001-3695.2014.11.073
基于协方差特征的裁剪AdaBoost算法
Pruning AdaBoost algorithm based on covariance feature
摘要
Abstract
With regard to the defect of Haar-like feature and the time-consuming training of AdaBoost algorithm,this paper presented an improved AdaBoost algorithm based on covariance feature.The method used covariance feature in place of Haar-like feature in feature extraction step,then combined the feature pruning and dynamic weight trimming in the training step.The experimental results show that,compared with AdaBoost algorithm based on Haar-like feature,the training time of this algo-rithm dramatically decreases without suffering a degeneration in classification capability.关键词
人脸检测/AdaBoost算法/协方差特征/特征裁剪/动态权重裁剪Key words
face detection/AdaBoost algorithm/covariance feature/feature pruning/dynamic weight trimming分类
信息技术与安全科学引用本文复制引用
李睿,李长风..基于协方差特征的裁剪AdaBoost算法[J].计算机应用研究,2014,(11):3517-3520,4.基金项目
国家自然科学基金资助项目 （）
甘肃省自然科学基金资助项目 （）
甘肃省财政厅科研项目 （）
