Presentation on theme: "Human Action Recognition by Learning Bases of Action Attributes and Parts Bangpeng Yao, Xiaoye Jiang, Aditya Khosla, Andy Lai Lin, Leonidas Guibas, and."— Presentation transcript:
1Human Action Recognition by Learning Bases of Action Attributes and Parts Bangpeng Yao, Xiaoye Jiang, Aditya Khosla, Andy Lai Lin, Leonidas Guibas, and Li Fei-FeiStanford University
2Outline Introduction Action Bases Learning the Dual-Sparse Action Bases and Reconstruction CoefficientsExperiments
3Introduction Human action recognition in still images Contributions A general image classification problemHuman-object interactionParts + AttributesContributionsRepresent each image by using a sparse set of action bases that are meaningful to the content of the imageEffectively learn these bases given far-from-perfect detections of action attributes and parts without meticulous human labeling
4Action Bases Attributes and parts Attributes: verb, learned by discriminative classifiersParts: object parts and poselets, learned by pre-trained object detectors and poselet detectorsA vector of the normalized confidence scores obtained from these classifiers and detectors is used to represent this image.
5Action Bases High-order interactions of image attributes and parts is used to represent each image and SVMs are trained for action classification