In this paper, we propose a novel transductive learning framework named manifold-ranking based image retrieval (MRBIR). Given a query image, MRBIR first makes use of a manifold ra...
Acquiring knowledge has long been the major bottleneck preventing the rapid spread of AI systems. Manual approaches are slow and costly. Machine-learning approaches have limitatio...
In this paper, we propose an approach that retrieves motion of objects from the videos based on the dynamic time warping of view invariant characteristics. The motion is represent...
Adaptivity, both of the individual agents and of the interaction structure among the agents, seems indispensable for scaling up multi-agent systems MAS's in noisy environme...
David Wolpert, Sergey Kirshner, Christopher J. Mer...
Abstract. We apply independent component analysis (ICA) for learning an efficient color image representation of natural scenes. In the spectra of single pixels, the algorithm was a...
Te-Won Lee, Thomas Wachtler, Terrence J. Sejnowski