Evolutionary Pruning for Fast and Robust Face Detection
Jun-Su Jang, Jong-Hwan Kim · 2006
Face detection task can be considered as a classifier training problem. It is a process to find the parameters of the classifier model by using the training data. To solve such a complex problem, evolutionary algorithm is employed in cascade structure of classifiers. In this paper, evolutionary pruning is proposed to reduce the number of weak classifiers in AdaBoost-based cascade detector while maintaining the detection accuracy. The computation time is proportional to the number of weak classifiers and therefore the reduction causes fast detection speed. The proposed cascade structure experimentally proves its efficient computation time. It is also compared with the state-of-the-art face detectors in terms of the detection accuracy, and the results show that the proposed method outperforms the previous studies.