Abstract | ||
---|---|---|
The present paper proposes a simultaneous estima- tion of both human head pose and shape. When head pose is estimated for a generic face model, a consid- erable estimation error is generated by the shape dif- ference between the model and the target face. In the proposed framework, a specific face model is gradually created for the target face by deforming the generic model based on an error analysis. These tasks are controlled by a hierarchical attention control system. Stable and efficient estimation is realized by switching the processes in a hierarchical control system. Exper- imental results show that the proposed system can si- multaneously estimate head pose and shape for several persons. mission lay-> |
Year | Venue | Keywords |
---|---|---|
2002 | MVA | attentional control |
Field | DocType | Citations |
Hierarchical control system,Computer vision,Pattern recognition,Computer science,3D pose estimation,Artificial intelligence,Human head | Conference | 0 |
PageRank | References | Authors |
0.34 | 2 | 2 |
Name | Order | Citations | PageRank |
---|---|---|---|
Junji Satake | 1 | 46 | 8.40 |
Takeshi Shakunaga | 2 | 192 | 43.46 |