Skip to main content
Figure 5 | EURASIP Journal on Audio, Speech, and Music Processing

Figure 5

From: A parameterizable spatiotemporal representation of popular dance styles for humanoid dancing characters

Figure 5

Process of key-pose synthesis with variability, from kinematic chains decomposition to the stochastic calculation of the joints' rotations. The top graph shows (a) the decomposition of the body model into five kinematic chains, and (b) the sequence of propagation of stochastic processes along the kinematic chain of the character's right arm. The bottom graph shows the proposed solution for generating the key-pose of metric class m with variability, by replicating the same variability observed in the recorded dance. This process can be implemented by (c) randomly calculating all key-pose's joint rotations: starting from the anchor segment, s 0 m , at the spine, which links joint 1 to joint 10, to the chain extremity at joint 20, each joint position, p j m , is randomly calculated inside its respective, C j m , by selecting a random quaternion, q v s j m , that describes a possible rotation of that joint segment, s j m , around its base unity vector, v s j m , (given by the last segment target vector, v s j m - 1 ), circumscribed by C j m .

Back to article page