Guest Editorial: Computational Vision at Brown

International Journal of Computer Vision -     

Covariant-Conics Decomposition of Quartics for 2D Shape Recognition and Alignment

This paper outlines a new geometric parameterization of 2D curves where parameterization is in terms of geometric invariants and parameters that determine intrinsic coordinate systems. This new approach handles two fundamental problems: single-computation alignment, and recognition of 2D shapes under Euclidean or affine transformations. The approach is model-based: every shape is first fitted by a quartic represented by a fourth degree 2D polynomial. Based on the decomposition of this equation into three covariant conics, we are able, in both the Euclidean and the affine cases, to define a unique intrinsic coordinate system for non-singular bounded quartics that incorporates usable alignment information contained in the polynomial representation, a complete set of geometric invariants, and thus an associated canonical form for a quartic. This representation permits shape recognition based on 11 Euclidean invariants, or 8 affine invariants. This is illustrated in experiments with real data sets.     

On Advances in Statistical Modeling of Natural Images

Statistical analysis of images reveals two interesting properties: (i) invariance of image statistics to scaling of images, and (ii) non-Gaussian behavior of image statistics, i.e. high kurtosis, heavy tails, and sharp central cusps. In this paper we review some recent results in statistical modeling of natural images that attempt to explain these patterns. Two categories of results are considered: (i) studies of probability models of images or image decompositions (such as Fourier or wavelet decompositions), and (ii) discoveries of underlying image manifolds while restricting to natural images. Applications of these models in areas such as texture analysis, image classification, compression, and denoising are also considered.     

On closed-loop adaptive noise cancellation

Given the mean limit ordinary differential equation for the stochastic approximation defining the adaptive algorithm for a closed-loop adaptive noise cancellation, we characterize the limit points. Under appropriate conditions, it is shown that as the dimension of the weight vector increases, the sequence of corresponding limit points converges in the sense of l₂ to the infinite-dimensional optimal weight vector. Also, the limit point of the algorithm is nearly optimal if the dimension of the weight vector is large enough. The gradient of the mean-square error with respect to the weight vector, evaluated at the limit, goes to zero in l₁ and l₂ as the dimension increases, as does the gradient with respect to the coefficients in the transfer function connecting the reference noise signal with the error output. Thus the algorithm is “nearly” a gradient descent algorithm and is error-reducing for large enough dimension. Under broad conditions, iterative averaging can be used to get a nearly optimal rate of convergence     

Hilbert-Schmidt lower bounds for estimators on matrix lie groupsfor ATR

Deformable template representations of observed imagery model the variability of target pose via the actions of the matrix Lie groups on rigid templates. In this paper, we study the construction of minimum mean squared error estimators on the special orthogonal group, SO(n), for pose estimation. Due to the nonflat geometry of SO(n), the standard Bayesian formulation of optimal estimators and their characteristics requires modifications. By utilizing Hilbert-Schmidt metric defined on GL(n), a larger group containing SO(n), a mean squared criterion is defined on SO(n). The Hilbert-Schmidt estimate (HSE) is defined to be a minimum mean squared error estimator, restricted to SO(n). The expected error associated with the HSE is shown to be a lower bound, called the Hilbert-Schmidt bound (HSB), on the error incurred by any other estimator. Analysis and algorithms are presented for evaluating the HSE and the HSB in cases of both ground-based and airborne targets     

Coping with Uncertainty in Map Learning

In many applications in mobile robotics, it is important for a robot to explore its environment in order to construct a representation of space useful for guiding movement. We refer to such a representation as a map, and the process of constructing a map from a set of measurements as map learning. In this paper, we develop a framework for describing map-learning problems in which the measurements taken by the robot are subject to known errors. We investigate approaches to learning maps under such conditions based on Valiant's probably approximately correct learning model. We focus on the problem of coping with accumulated error in combining local measurements to make global inferences. In one approach, the effects of accumulated error are eliminated by the use of local sensing methods that never mislead but occasionally fail to produce an answer. In another approach, the effects of accumulated error are reduced to acceptable levels by repeated exploration of the area to be learned. We also suggest some insights into why certain existing techniques for map learning perform as well as they do. The learning problems explored in this paper are quite different from most of the classification and boolean-function learning problems appearing in the literature. The methods described, while specific to map learning, suggest directions to take in tackling other learning problems.     

Installation and commissioning of the CDMSII experiment at Soudan

In the past year and a half, the Cryogenic Dark Matter Search (CDMS) collaboration has been active at the Soudan mine in installing a system for running ZIP detectors that will be used to search for dark matter in the form of Weakly Interacting Massive Particles. Presently, there is an operating cryogenic system, working electronics, a functional data acquisition and analysis system, passive shielding, an active muon veto, and 12 ZIP detectors. Six of the 12 ZIP detectors have been tested in situ and are fully operational with acceptable noise profiles. CDMS is in the process of commissioning the experiment and expects to be making a background measurement by the end of summer 2003.     

Study of the dead layer in germanium for the CDMS detectors

We present new measurements on a Cryogenic Dark Matter Search (CDMS) detector with electron, neutron, and gamma sources. The measurements have been performed to investigate the dead layer of one of the CDMS Z-dependent Ionization Phonon germanium detectors. The dead layer has been studied at both charge electrodes and at different electric field intensities. We also present a method to remove the dependence of athermal phonon measurements on event position.     

Coarsening of boron carbide grains during the infiltration of porous boron carbide preforms by molten silicon

This work investigates the coarsening of boron carbide grains during the infiltration of porous boron carbide preforms by molten silicon with respect to fabrication of reaction-bonded boron carbide ceramics. Experimental results reveal that the shape of boron carbide grains evolve from the irregular shape to faceted shape due to dissolution-precipitation during infiltration. For infiltration temperatures below 1750 °C, the boron carbide grains are irregular and exhibit an unimodal size distribution, which can be ascribed to the normal grain growth. The growth of the irregular grains follow a cubic law of diffusion control. In contrast, for infiltration temperatures above 1750 °C, the boron carbide grains become faceted and exhibit a bimodal size distribution, indicative of the typical abnormal grain growth. The abnormal growth of faceted grains is proposed to be controlled by coalescence-enhanced two-dimensional nucleation.     

Don't Judge a (Face)Book by its Cover: Exploring judgement accuracy of others' personality on Facebook

With interaction across social networking sites (SNSs) becoming more prominent, the ability to accurately judge another's personality through these digital platforms is an important area of investigation. A number of studies demonstrate that SNSs can be an effective means of communicating information on personality (Evans, Gosling & Carroll, 2008). Much less research has examined the online cues that people use when forming their impressions of another's personality. The current research adopted a mixed methods approach to examine 1) accuracy of first impressions formed on the basis of viewing a person's Facebook behaviour and 2) how judgements are formed. Findings suggested that the traits of openness and conscientiousness can be judged most accurately through observing Facebook behaviour. To explore the cues used in forming personality judgements, thematic analysis was undertaken. This revealed six information ‘cues’ including; (1) vocabulary of target, (2) photographs, (3) written online interactions, (4) relationships with others, (5) health status and, (6) occupational status. Findings are discussed in terms of Funder's Realistic Accuracy Model (1995, 1999) along with limitations and suggestions for future research.