the ball, gym movements (like jump, twist and squat) and running-leaping. 3D human motions are visually realistic and skin is smooth and natural even in highly articulated dancing poses. We also quantitatively compare Method-I and the proposed pipeline using Eqn. 1. Results are shown in Table. 4. It is obvious that our method has smaller kinematic error and provides better accuracy. Additionally, we test our pipeline on different mocap data sets, including data from Ohio State University mocap lab [2] and ViHASi [25]. Our method can generate natural motions using both data sets and various human shape models. In Fig. 10, we show a few animations from two mocap data sets using four different shape models.