Charles R. Dyer

Bio: Charles R. Dyer is an academic researcher from University of Wisconsin-Madison. The author has contributed to research in topics: Motion estimation & Motion field. The author has an hindex of 43, co-authored 141 publications receiving 9919 citations. Previous affiliations of Charles R. Dyer include University of Wisconsin System & University of Maryland, College Park.

A cognitive architecture for computer vision

Abstract: Our objective is to design a visual analyzer for real-time scene analysis of dynamic imagery. For this purpose we employ a cognitive architecture derived from concepts of the central nervous system. Output of the visual analyzer will be high-level real-time perceptual data suitable either for the guidance of robot action or for the symbolic description of the visual environment. This output resides in a distributed relational database modeling the hypercolumnar architecture of the visual cortex.

Efficient Schemes for Monte Carlo Markov Chain Algorithms in Global Illumination

Abstract: Current MCMC algorithms are limited from achieving high rendering efficiency due to possibly high failure rates in caustics perturbations and stratified exploration of the image plane. In this paper we improve the MCMC approach significantly by introducing new lens perturbation and new path-generation methods. The new lens perturbation method simplifies the computation and control of caustics perturbation and can increase the perturbation success rate. The new path-generation methods aim to concentrate more computation on "high perceptual variance" regions and "hard-to-find-but-important" paths. We implement these schemes in the Population Monte Carlo Energy Redistribution framework to demonstrate the effectiveness of these improvements. In addition., we discuss how to add these new schemes into the Energy Redistribution Path Tracing and Metropolis Light Transport algorithms. Our results show that rendering efficiency is improved with these new schemes.

Object Exploration By Purposive, Dynamic Viewpoint Adjustment

Abstract: We present a viewing strategy for exploring the surface of an unknown object (i.e., making all of its points visible) by purposefully controlling the motion of an active observer. It is based on a simple relation between (1) the instantaneous direction of motion of the observer, (2) the visibility of points projecting to the occluding contour, and (3) the surface normal at those points: If the dot product of the surface normal at such points and the observer's velocity is positive, the visibility of the points is guaranteed under an innnitesi-mal viewpoint change. We show that this leads to an object exploration strategy in which the observer purposefully controls its motion based on the occluding contour in order to impose structure on the set of surface points explored, make its representation simple and qualitative, and provably solve the exploration problem for smooth generic surfaces of arbitrary shape. Unlike previous approaches where exploration is cast as a discrete process (i.e., asking where to look next?) and where the successful exploration of arbitrary objects is not guaranteed, our approach demonstrates that dynamic viewpoint control through directed observer motion leads to a qualitative exploration strategy that is provably-correct, depends only on the dynamic appearance of the occluding contour, and does not require the recovery of detailed three-dimensional shape descriptions from every position of the observer. The support of the National Science Foundation under grants IRI-9022608 and IRI-9196106 is gratefully acknowledged.

Real-time, parallel motion tracking of three dimensional objects from spatiotemporal

Abstract: A major issue in computer vision is the interpretation of three-dimensional (3D) motion of moving objects from a continuous stream of two-dimensional (2D) images. In this paper we consider the problem where the 3D motion of an object corresponding to a known 3D model is to be tracked using only the motion of 2D features in the stream of images. Two general solution paradigms for this problem are characterized: (1) motion-searching, which hypothesizes and tests 3D motion parameters, and (2) motion-calculating, which uses back­ projection to directly estimate 3D motion from image-feature motion. Two new algorithms for computing 3D motion based on these two paradigms are presented. One of the major novel aspects of both algorithms is their use of the assumption that the input image stream is spatiotemporally dense. This constraint is psychologically plausible since it is also used by the short-range motion processes in the human visual system. The processing of a temporally-unbounded, spatiotemporal image com­ bined with the resource constraint of finite image buffer memory, requires real-time throughput rates for our algorithms. Consequently, another major focus of this paper is the development of real-time, parallel implementations to achieve the required throughput. Implementations of both algorithms are described using an Aspex Pipe for low-level, image-feature computations and a Sequent Symmetry for high-level, model-based computations. The Pipe, a pipelined image processor, is tightly-coupled with the Sequent, and semaphores are used for synchronization between the two. Design issues and parallel im­ plementation issues of both algorithms are discussed in detail.

A method for registration of 3-D shapes

Abstract: The authors describe a general-purpose, representation-independent method for the accurate and computationally efficient registration of 3-D shapes including free-form curves and surfaces. The method handles the full six degrees of freedom and is based on the iterative closest point (ICP) algorithm, which requires only a procedure to find the closest point on a geometric entity to a given point. The ICP algorithm always converges monotonically to the nearest local minimum of a mean-square distance metric, and the rate of convergence is rapid during the first few iterations. Therefore, given an adequate set of initial rotations and translations for a particular class of objects with a certain level of 'shape complexity', one can globally minimize the mean-square distance metric over all six degrees of freedom by testing each initial registration. One important application of this method is to register sensed data from unfixtured rigid objects with an ideal geometric model, prior to shape inspection. Experimental results show the capabilities of the registration algorithm on point sets, curves, and surfaces. >

Object Detection with Discriminatively Trained Part-Based Models

Abstract: We describe an object detection system based on mixtures of multiscale deformable part models. Our system is able to represent highly variable object classes and achieves state-of-the-art results in the PASCAL object detection challenges. While deformable part models have become quite popular, their value had not been demonstrated on difficult benchmarks such as the PASCAL data sets. Our system relies on new methods for discriminative training with partially labeled data. We combine a margin-sensitive approach for data-mining hard negative examples with a formalism we call latent SVM. A latent SVM is a reformulation of MI--SVM in terms of latent variables. A latent SVM is semiconvex, and the training problem becomes convex once latent information is specified for the positive examples. This leads to an iterative training algorithm that alternates between fixing latent values for positive examples and optimizing the latent SVM objective function.

A taxonomy and evaluation of dense two-frame stereo correspondence algorithms

Abstract: Stereo matching is one of the most active research areas in computer vision. While a large number of algorithms for stereo correspondence have been developed, relatively little work has been done on characterizing their performance. In this paper, we present a taxonomy of dense, two-frame stereo methods designed to assess the different components and design decisions made in individual stereo algorithms. Using this taxonomy, we compare existing stereo methods and present experiments evaluating the performance of many different variants. In order to establish a common software platform and a collection of data sets for easy evaluation, we have designed a stand-alone, flexible C++ implementation that enables the evaluation of individual components and that can be easily extended to include new algorithms. We have also produced several new multiframe stereo data sets with ground truth, and are making both the code and data sets available on the Web.

Planning Algorithms: Introductory Material

Abstract: Planning algorithms are impacting technical disciplines and industries around the world, including robotics, computer-aided design, manufacturing, computer graphics, aerospace applications, drug design, and protein folding. This coherent and comprehensive book unifies material from several sources, including robotics, control theory, artificial intelligence, and algorithms. The treatment is centered on robot motion planning but integrates material on planning in discrete spaces. A major part of the book is devoted to planning under uncertainty, including decision theory, Markov decision processes, and information spaces, which are the “configuration spaces” of all sensor-based planning problems. The last part of the book delves into planning under differential constraints that arise when automating the motions of virtually any mechanical system. Developed from courses taught by the author, the book is intended for students, engineers, and researchers in robotics, artificial intelligence, and control theory as well as computer graphics, algorithms, and computational biology.