A Fast Sub-Volume Search Method for Human Action Detection

作者：Guo Ping^*; Miao Zhenjiang; Zhang Xiao Ping; Wang Zhe

来源：IEICE Transactions on Information and Systems, 2012, E95D(1): 285-288.

DOI：10.1587/transinf.E95.D.285

• action detection
• graph representation
• Minimum Cycle detection
• sub-volume search

This paper discusses the task of human action detection. It requires not only classifying what type the action of interest is, but also finding actions%26apos; spatial-temporal locations in a video. The novelty of this paper lies on two significant aspects. One is to introduce a new graph based representation for the search space in a video. The other is to propose a novel sub-volume search method by Minimum Cycle detection. The proposed method has a low computation complexity while maintaining a high action detection accuracy. It is evaluated on two challenging datasets which are captured in cluttered backgrounds. The proposed approach outperforms other state-of-the-art methods in most situations in terms of both Precision-Recall values and running speeds.

• 出版日期2012-1
• 单位北京交通大学