Training more discriminative multi-class classifiers for hand detection
作者：
Highlights：
• A set of shared stumps are combined to strengthen the discrimination power of weak classifiers.
• A “slowest error growth” discriminant to determine the optimal combination of stumps.
• Multiple thresholds are leveraged in shared stumps to fit different classes.
• We associate effective features with different classes of hands, and employ mix-type features.
• As compared with JointBoost, our classifier can obtain a better classification performance with less runtime cost.
摘要
Highlights•A set of shared stumps are combined to strengthen the discrimination power of weak classifiers.•A “slowest error growth” discriminant to determine the optimal combination of stumps.•Multiple thresholds are leveraged in shared stumps to fit different classes.•We associate effective features with different classes of hands, and employ mix-type features.•As compared with JointBoost, our classifier can obtain a better classification performance with less runtime cost.
论文关键词：Multi-class classifiers,Hand detection,Classifier combination,Boosting,Stump classifiers
论文评审过程：Received 3 May 2014, Revised 16 July 2014, Accepted 2 September 2014, Available online 16 September 2014.
论文官网地址：https://doi.org/10.1016/j.patcog.2014.09.001
