Many researchers have taken the effort to describe the dynamics of the articulated body by the analytic method. They have obtained excellent results in various fields. However, for the articulated body moving with its voluntary will, it is difficult to generalize the motion pattern by analytical modeling, because the motion pattern is extremely subjective and unpredictable. The learning networks overcome the restriction of analytic modeling through the deductive learning method. The Uniform Posture Map (UPM) is proposed to synthesize a new motion between existing clip motions. It is organized through the quantization of various postures with an unsupervised learning algorithm; it places the output neurons with similar postures in adjacent positions. Using this property, an intermediate posture of applied two postures is generated; the generating posture is used as a key-frame to make an interpolating motion. The UPM needs fewer computational costs, in comparison with other motion transition algorithms. It provides a control parameter; an animator can not only control the motion simply by adjusting this parameter, but also produce animation interactively. The UPM prevents the generating of the invalid output neurons to present unreal postures in the learning phase; thus, it makes more realistic motion curves; finally it contributes to the making of more natural motions.
展开▼
