We have achieved combined model acquisition and control by a combination of techniques:
A combined model acquisition and control approach has many advantages. In addition to permitting uncalibrated visual servo control, the on-line estimated models are useful for (1) prediction and constraining search in visual tracking (Section 4.1 and [Jägersand and Nelson, 1995, Feddema and Lee, 1990]), (2) performing local coordinate transformations between manipulator (joint), world, and visual frames (Chapter 7 and [Jägersand and Nelson, 1994, Jägersand and Nelson, 1995]), and (3) synthesizing views from a basis of agent poses (Chapter 6 and [Jägersand, 1997]). We have found such an adaptive approach to be particularly helpful in robot arm manipulation when carrying out difficult tasks, such as manipulation of flexible material Section 5.9 and [Jägersand and Nelson, 1994, Jägersand and Nelson, 1995]), or performing large rotations for exploring object shape [Kutulakos and Jägersand, 1995]. For a dextrous multi-finger robot hand, such as the Utah/MIT hand, the fully adaptive approach is appealing because dextrous manipulation of a grasped object is much harder to model accurately than a typical robot arm system, where the object is rigidly attached to the end effector (Section 5.10 and [Jägersand et al., 1996]).