Evaluating spatiotemporal interest point features for depth-based action recognition

• Published in: Image and vision computing Vol. 32; no. 8; pp. 453 - 464
• Main Authors: Zhu, Yu, Chen, Wenbin, Guo, Guodong
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 01.08.2014
• Subjects: Action recognition; Descriptors; Detectors; Evaluation; Feature fusion; Feature recognition; Human; Imaging; Recognition; RGB-D sensor; Spatiotemporal interest point (STIP); State of the art; STIP feature refinement; STIP features; Support vector machines; Three dimensional; Action recognition; STIP feature refinement; Evaluation; RGB-D sensor; Detectors; Feature fusion; Descriptors; Spatiotemporal interest point (STIP); STIP features

Human action recognition has lots of real-world applications, such as natural user interface, virtual reality, intelligent surveillance, and gaming. However, it is still a very challenging problem. In action recognition using the visible light videos, the spatiotemporal interest point (STIP) based f...