Paul R. Cooper

Arc consistency: parallelism and domain dependence

Abstract This paper discusses how better arc consistency algorithms for constraint satisfaction can be developed by exploiting parallelism and domain-specific problem characteristics. A massively parallel algorithm for arc consistency is given, expressed as a digital circuit. For a constraint satisfaction problem with n variables and a labels, this algorithm has a worst-case time complexity of O(na), significantly better than that of the optimal uniprocessor algorithm. An algorithm of intermediate parallelism suitable for implementation on a SIMD machine is also given. Analyses and implementation experiments are shown for both algorithms. A method for exploiting characteristics of a problem domain to achieve arc consistency algorithms with better time and space complexity is also discussed. A general technique for expressing domain knowledge and using it to develop optimized arc consistency algorithms is described. The domain-specific optimizations can be applied analogously to any of the arc consistency algorithms along the sequential/parallel spectrum.

STRUCTURE RECOGNITION BY CONNECTIONIST RELAXATION: FORMAL ANALYSIS

A formal description is given of a connectionist implementation of discrete relaxation for labelled graph matching. The network is shown to converge. The desired behavior of the algorithm is formally specified; then it is proved that the result of the relaxation meets the formal goal. The network is limited by complexity considerations to the detection and propagation of unary and binary consistency constraints. The application is fast parallel indexing into a memory of object models, based on a visually derived junction/link structure description. Implementation experiments are presented, and explicit and exact space and time requirements are developed.

Looking for trouble: Using causal semantics to direct focus of attention

Vision should provide an explanation of the scene in terms of a causal semantics. The authors propose a characterization of what constitutes visual understanding. The cornerstone of the proposal is that visual understanding is fundamentally a matter of developing a causal explanation of the scene, i.e., of determining the causal significance of the elements in a scene, and the causal relationships among those elements. Simple, naive physical knowledge is used as the basis of a vertically integrated vision system that explains arbitrarily complex stacked block structures. The semantics provides a basis for controlling the application of visual attention, and forms a framework for the explanation that is generated. It is shown that the program sequentially explores scenes of complex blocks structures, identifies functional substructures such as arches and cantilevers, and develops an explanation of why the whole construction stands and the role of each block in its stability.<<ETX>>

The dynamic retina: contrast and motion detection for active vision

This paper presents an efficient, biologically-inspired early vision architecture, the dynamic retina, that is well-suited to highly active and responsive vision platforms. The dynamic retina exploits normally undesirable camera motion as a necessary step in detecting image contrast, by using dynamic receptive fields instead of traditional spatial-neighborhood operators. We analyze the continuous miniature “noise” movements made by active imaging systems, and show that they can be exploited to detect contrast. We then develop an appropriate photoreceptor response function, based on light-adaptation models for vertebrate receptors. Together, the movements and response function over time compute image contrast. The dynamic retina is also useful for motion analysis, since moving objects processed by the system leave a clear signature from which motion parameters can be extracted. Results from a number of experiments with real video sequences demonstrate the effectiveness of the system for both contrast detection and motion analysis.

Causal scene understanding

Abstract Most computer vision systems are concerned with computing the whats and wheres of a scene. We describe a set of programs concerned instead with computing the whys and hows—why the scene is the way it is, and how an agent can interact with it. The basis of our approach lies in the construction of a causal explanation of a scene—a representation that describes what affects what in the scene, how these elements affect each other, and why they affect each other the way they do. Such explanations, by definition and design, must encompass representations of the potentials for action in a scene, and thus form a natural basis for describing how scene elements serve purposes—i.e., functional descriptions. As a concrete case study in causal scene understanding, this paper focuses primarily on ways to exploit the causality of objects in static equilibrium, in particular, the causality of support. We describe three camera-to-commentary vision systems, operating in three different domains, that develop causal explanations of scenes from visual images of those scenes and, in the process, provide novel solutions to a number of traditional problems in vision and robotics, including occlusion, focus of attention, and grasp planning. We also show how the kinds of causal descriptions produced by these systems can be exploited to physically interact with the scene.

Parallel structure recognition with uncertainty: coupled segmentation and matching

A network is described that recognizes objects from uncertain image-derivable descriptions. The network handles uncertainty by making the recognition and segmentation decisions simultaneously, in a cooperative way. Both problems are posed as labeling problems, and a coupled Markov random field (MRF) is used to provide a single formal framework for both. Prior domain knowledge is represented as weights within the MRF network and interacts with the evidence to yield a labeling decision. The domain problem is the recognition of structured objects composed of simple junction and link primitives. Implementation experiments demonstrate the parallel segmentation and recognition of multiple objects in noisy ambiguous scenes with occlusion.<<ETX>>

A Markov random field model of subjective contour perception

We describe a model of subjective contour perception that sees the triangle: top-down feedback from the perception of an occluding object causes the perception of illusory contrast along the objects boundary. Unlike other approaches, the model explicitly incorporates a representational hierarchy of visual elements at multiple levels of abstraction, for example, both local contrast edges and triangles. The model is a connectionist network implemented within the formal framework of Markov random fields, which provides for bidirectional information flow over connections. Simulation results show how feedforward activation yields the perception of the occluding triangle at a high-level of abstraction, and how feedback activates the units that represent the illusory contrast.

Early Vision Network for a Moving Eye: Dynamic Contrast and Motion Detection

ABSTRACT We present a biologically-inspired early vision network that is well-suited to highly active and responsive visionplatforms. The network exploits normally undesirable camera motion as a necessary step in detecting imagecontrast. It also detects visual motion, producing distinctive signals from which useful image motion parametersare extracted. The network remains sensitive over a very wide dynamic range of inputs, and has self-calibratingproperties that make it amenable to analog VLSI implementation. The results also support the hypothesis that vertebrate cones function primarily as detectors of contrast and motion, rather than intensity. Experiments verify that naturally occurring jitter in a motor-mounted camera, instead of being avoided, can be exploited inearly visual processing. 1 INTRODUCTION Active vision systems such as mammalian eyes or mobile cameras are in constant motion. Even when fixated forapproximately stationary image acquisition, the sensing platform is rarely completely still. For active vision systems,it is impractical to hold the camera still while sensing.This paper describes a system for early visual processing, motivated by vertebrate retinal processing, that ezploitsthis unavoidable motion normally regarded as noise. The effects of physical camera jitter, instead of being removed