Autonomous star recognition requires that many similar patterns must be distinguished from one another with a small training set. Since these systems are implemented on-board a spacecraft, the network needs to have low memory requirements and minimal computational complexity. Fast training speeds are also important since star sensor capabilities change over time. This paper compares two networks that meet these needs: radial basis function networks and neural logic networks. Neural logic networks performed much better than radial basis function networks in terms of recognition accuracy, memory needed, and training speed.

The existence of two- and three-finger grasps in the presence of arbitrarily small friction is shown for 2D and 3D smooth objects using a simple technique. No convexity of the objects is assumed. The existence of finger gaits for rotating a planar object using three and four fingers is proved. Additional results for smooth convex objects which describe two different finger gaits using four fingers are presented. The purpose is to move one or more fingers at any one time and at the same time to maintain a grip with the others, and thereby to move the object further than or in other ways than it could be moved with a fixed grip.>