A discriminative representation for human action recognition

作者：

Highlights：

• A discriminative representation is proposed by discovering key information of the input data.

• The representation is parameterized with hidden variables and can be learned from training data.

• Human action is recognized by combining the parameterized representation and discriminative classifier.

• The performance of action recognition is improved by updating the representation and the classifier alternatively.

摘要

Highlights•A discriminative representation is proposed by discovering key information of the input data.•The representation is parameterized with hidden variables and can be learned from training data.•Human action is recognized by combining the parameterized representation and discriminative classifier.•The performance of action recognition is improved by updating the representation and the classifier alternatively.

论文关键词：Action recognition,Discriminative representation,Classifier,Maximum likelihood

论文评审过程：Received 11 September 2015, Revised 21 February 2016, Accepted 27 February 2016, Available online 10 March 2016, Version of Record 23 August 2016.

论文官网地址：https://doi.org/10.1016/j.patcog.2016.02.022