Research on combined asymmetric AdaBoost for face detection

Face detection is a widely studied topic in computer vision. Despite of its great success, several key problems are still unresolved for AdaBoost algorithms: how to select the asymmetric weak learners and how to combine added sample sets.In this paper,a new combined asymmetric AdaBoost algotithms is proposed to make improvement in the two aspects.First we select the asymmetric weak learners by computering sample distribution of positive sample and negative sample. Second,we combine the added sample sets by boosting chain. Last, we have used this new method for face detection. Experiments with synthetic and real scene data sets show our algorithm outperforms conventional AdaBoost.