The problem of learning is arguably at the very core of the problem of intelligence, both biological and artificial. In this paper we review our work over the last ten years in the area of supervised learning, focusing on three interlinked directions of research: theory, engineering applications (making intelligent software) and neuroscience (understanding the brain’s mechanisms of learning) wh...

The study of visual object recognition is often motivated by the problem of recognizing 3-d objects given that we receive 2-d patterns of light on our retinae. Recent findings from human psychophysics, neurophysiology and computational vision provide converging evidence for a view-based recognition framework in which objects and scenes are represented as collections of viewpoint-specific local ...

In the context of part segmentation from 2D edge images, this paper presents some interesting results with a novel method that addresses the problem of ltering a redundant set of part hypotheses that retains only those that are likely to correspond to actual parts. In the proposed method, supporting evidence for hypotheses are put in competition in a Minimum Description-Length (MDL) framework t...

Abstract A method is developed to track planar and nearplanar objects by incorporating a model of the expected image template distortion, and fitting the sampling region to pre-trained examples with general regression. The approach does not assume a particular form of the underlying space, allows a natural handling of occluding objects, and permits dynamic changes of the scale and size of the s...

This paper surveys the field of machine vision from a computer science perspective. It is written to act as an introduction to the field and presents the reader with references to specific implementations. Machine vision is a complex and developing field that can be broken into the three stages: stereo correspondence, scene reconstruction, and object recognition. We present the techniques and g...

The tracking of objects in real time lends itself to vast industrial and practical applications; of interest is the tracking of objects within a manipulator's workspace by using machine vision. This paper presents the basics for control by vision of a 3-DOF robotic manipulator using artificial intelligence. The object or objects are first identified, next the object's movement or trajectory is ...

The wide-spread use of digital cameras has already created large markets for efficient digital image processing and compression techniques. Regardless of this, the great majority of current camera systems are still blind, lacking the ability to understand the contents of the imagery taken by the cameras. The goal of computer vision (machine vision) is to make useful decisions about real physica...