Trajectory-based human action segmentation

This paper proposes a sliding window approach, whose length and time shift are dynamically adaptable in order to improve model confidence, speed and segmentation accuracy in human action sequences. Activity recognition is the process of inferring an action class from a set of observations acquired b...

Full description

Bibliographic Details
Main Author: Santos, Luís (author)
Other Authors: Khoshhal, Kamrad (author), Dias, Jorge (author)
Format: article
Language:eng
Published: 2015
Subjects:
Motion segmentation
Classification framework
Signal processing
Motion variability
Adaptive sliding window
Online Access:http://hdl.handle.net/10316/27816
Country:Portugal
Oai:oai:estudogeral.sib.uc.pt:10316/27816

Internet

http://hdl.handle.net/10316/27816

Similar Items