The Time Dimension for Scene Analysis

Abstract

A fundamental issue in neural computation is the binding problem, which refers to how sensory elements in a scene organize into perceived objects, or percepts. The issue of binding is hotly debated in recent years in neuroscience and related communities. Much of the debate, however, gives little attention to computational considerations. This review intends to elucidate the computational issues that bear directly on the binding issue. The review starts with two problems considered by Rosenblatt to be the most challenging to the development of perceptron theory more than 40 years ago, and argues that the main challenge is the figure-ground separation problem, which is intrinsically related to the binding problem. The theme of the review is that the time dimension is essential for systematically attacking Rosenblatt's challenge. The temporal correlation theory as well as its special form--oscillatory correlation theory-is discussed as an adequate representation theory to address the binding problem. Recent advances in understanding oscillatory dynamics are reviewed, and these advances have overcome key computational obstacles for the development of the oscillatory correlation theory. We survey a variety of studies that address the scene analysis problem. The results of these studies have substantially advanced the capability of neural networks for figure-ground separation. A number of issues regarding oscillatory correlation are considered and clarified. Finally, the time dimension is argued to be necessary for versatile computing.

A Stereovision Matching Strategy for Images Captured with Fish-Eye Lenses in Forest Environments

“…To deal with real-world scenes some criterion for grouping elements in the scene is required. In the work of [28] a list of major grouping principles is exhaustively studied. They are inspired in the Gestalt’s principles [29].…”

Oscillatory correlation for scene segmentation of hyperspectral imagery for geospatial analysis

“…12,13 This is analogous to how people can only pay attention to a limited number of objects at a time. It is plausible to think that early and midlevel visual analysis provides region candidates for selection in a higherlevel visual process that can only track a limited number of things at time.…”

“…Wang). 12 In this study, we investigate a multiband version of the LEGION algorithm capable of processing hyperspectral imagery. This multiband LEGION is implemented by modifying the connection weights in the network such that closeness is measured by the Euclidean distance between pixel vectors in a local neighborhood.…”

A Network of Neural Oscillators for Fractal Pattern Recognition

“…Developing artificial pattern recognition systems as well as investigating actual natural mechanisms of pattern recognition in humans and animals has been a scientific challenge for decades [6,[9][10][11][12][13]. Major contributions to the dynamics of neural networks were made in the early 1970s and 1980s by Grossberg [9] and Hopfield [14].…”