Study on semantic control in motion synthesis

This paper focuses on studying semantic control in motion graph based human motion synthesis. Relational features are implemented for describing and specifying motions with spatial variations within a class of semantically similar motions at a high semantic level. Motion templates are then automatically derived from the training motions for capturing the spatio-temporal characteristics of an entire given class of semantically related motions. Automatic semantic annotation is performed on the unknown motion data document by identifying the presence of certain motion class from matching their respective motion class templates. Finally, the semantic control is introduced in motion graph based human motion synthesis. Experiments of motion synthesis demonstrate the effectiveness of the approach which enables user the higher level of intuitive semantically control and quality of human motion synthesis from mocap database.