Title | ||
---|---|---|
Clustering of human actions using invariant body shape descriptor and dynamic time warping |
Abstract | ||
---|---|---|
We propose a human action clustering method based on a 3D representation of the body in terms of volumetric coor- dinates. Features representing body postures are extracted directly from 3D data, making the system inherently insensi- tive to viewpoint dependence, motion ambiguities and self- occlusions. An Invariant Shape Descriptor of human body is obtained in order to capture only posture-dependent char- acteristics, despite possible differences in translation, ori- entation, scale and body size. Frame-by-frame descriptions, generated from a gesture sequence, are collected together in matrices. Clustering of action matrices is eventually performed, and through a Dynamic Time Warping (while computing the distance metric), we gain independence from possible temporal nonlinear distortions among different in- stances of the same gesture. |
Year | DOI | Venue |
---|---|---|
2005 | 10.1109/AVSS.2005.1577237 | AVSS |
Keywords | Field | DocType |
gesture recognition,image segmentation,nonlinear distortion,pattern clustering,action matrices,dynamic time warping,frame-by-frame descriptions,gesture sequence,human actions clustering,invariant body shape descriptor,posture-dependent characteristics,temporal nonlinear distortions,volumetric coordinates | Computer vision,Dynamic time warping,Pattern recognition,Computer science,Gesture,Matrix (mathematics),Metric (mathematics),Gesture recognition,Image segmentation,Invariant (mathematics),Artificial intelligence,Cluster analysis | Conference |
Citations | PageRank | References |
14 | 0.70 | 12 |
Authors | ||
4 |
Name | Order | Citations | PageRank |
---|---|---|---|
Massimiliano Pierobon | 1 | 585 | 49.21 |
Marco Marcon | 2 | 52 | 11.10 |
Augusto Sarti | 3 | 462 | 81.26 |
Stefano Tubaro | 4 | 1033 | 119.50 |