Alistair Sutherland

Bio: Alistair Sutherland is an academic researcher from Dublin City University. The author has contributed to research in topics: Gesture recognition & Feature extraction. The author has an hindex of 16, co-authored 70 publications receiving 1763 citations. Previous affiliations of Alistair Sutherland include Hitachi & University of Strathclyde.

A hand gesture recognition system and method

Abstract: Noise problems in processing small images or large-granularity images are reduced by representing hand images as rotational vectors calculated using a real-valued centroid. The hand region is therefore sectored independently of pixel quantization. Color segmentation is used to identify hand-color regions, followed by region labelling to filter out noise regions based on region size. Principal component analysis is used to plot gesture models.

Statlog: comparison of classification algorithms on large real-world problems

Abstract: This paper describes work in the StatLog project comparing classification algorithms on large real-world problems The algorithms compared were from symbolic learning (CART C45, NewID, AC2,ITrule, Cal5, CN2), statistics (Naive Bayes, k-nearest neighbor, kernel density, linear discriminant, quadratic discriminant, logistic regression, projection pursuit, Bayesian networks), and neural networks (backpropagation, radial basis functions) Twelve datasets were used: five from image analysis, three from medicine, and two each from engineering and finance We found that which algorithm performed best depended critically on the data set investigated We therefore developed a set of data set descriptors to help decide which algorithms are suited to particular data sets For example, data sets with extreme distributions (skew > l and kurtosis > 7) and with many binary/categorical attributes (>38%) tend to favor symbolic learning algorithms We suggest how classification algorithms can be extended in a number of d

Facial image processing

Abstract: A camera (10) captures a subject facial image. A unit (30) generates a tracking signal containing feature extraction data. In conjunction with a memory store (40), the unit (30) generates an output signal representing a substitute face having the features of the subject face. To generate the tracking signal, the facial area is detected (3103, 3107-8) initially and used in location of the mouth (3102, 3110-3114) and eyes (3104-6, 3115-3118). Only the H and S components are used for detecting the facial area; and only the S and V components for the mouth within the facial area. A face vector may be generated (50) using the tracking signals.

Finding Spiral Structures in Images of Galaxies

Abstract: Much recent work in statistical image analysis has been concerned with `cleaning' images by a bayesian statistical analysis incorporating a prior model, which reflects the spatial structure of the image. In almost all cases this has involved a description of the image at pixel level. In this paper we take the process further, and develop a spatial stochastic process of objects present in the image. The general theory is given and applied to images of spiral galaxies, with the aims of producing better schematic reconstructions and of automatically classifying galaxies.

Sequential Monte Carlo methods in practice

Abstract: Monte Carlo methods are revolutionizing the on-line analysis of data in fields as diverse as financial modeling, target tracking and computer vision. These methods, appearing under the names of bootstrap filters, condensation, optimal Monte Carlo filters, particle filters and survival of the fittest, have made it possible to solve numerically many complex, non-standard problems that were previously intractable. This book presents the first comprehensive treatment of these techniques, including convergence results and applications to tracking, guidance, automated target recognition, aircraft navigation, robot navigation, econometrics, financial modeling, neural networks, optimal control, optimal filtering, communications, reinforcement learning, signal enhancement, model averaging and selection, computer vision, semiconductor design, population biology, dynamic Bayesian networks, and time series analysis. This will be of great value to students, researchers and practitioners, who have some basic knowledge of probability. Arnaud Doucet received the Ph. D. degree from the University of Paris-XI Orsay in 1997. From 1998 to 2000, he conducted research at the Signal Processing Group of Cambridge University, UK. He is currently an assistant professor at the Department of Electrical Engineering of Melbourne University, Australia. His research interests include Bayesian statistics, dynamic models and Monte Carlo methods. Nando de Freitas obtained a Ph.D. degree in information engineering from Cambridge University in 1999. He is presently a research associate with the artificial intelligence group of the University of California at Berkeley. His main research interests are in Bayesian statistics and the application of on-line and batch Monte Carlo methods to machine learning. Neil Gordon obtained a Ph.D. in Statistics from Imperial College, University of London in 1993. He is with the Pattern and Information Processing group at the Defence Evaluation and Research Agency in the United Kingdom. His research interests are in time series, statistical data analysis, and pattern recognition with a particular emphasis on target tracking and missile guidance.

C ONDENSATION —Conditional Density Propagation forVisual Tracking

Abstract: The problem of tracking curves in dense visual clutter is challenging. Kalman filtering is inadequate because it is based on Gaussian densities which, being unimo dal, cannot represent simultaneous alternative hypotheses. The Condensation algorithm uses “factored sampling”, previously applied to the interpretation of static images, in which the probability distribution of possible interpretations is represented by a randomly generated set. Condensation uses learned dynamical models, together with visual observations, to propagate the random set over time. The result is highly robust tracking of agile motion. Notwithstanding the use of stochastic methods, the algorithm runs in near real-time.

Gestures for touch sensitive input devices

Abstract: Methods and systems for processing touch inputs are disclosed. The invention in one respect includes reading data from a multipoint sensing device such as a multipoint touch screen where the data pertains to touch input with respect to the multipoint sensing device, and identifying at least one multipoint gesture based on the data from the multipoint sensing device.

Method and apparatus for integrating manual input

Abstract: Apparatus and methods are disclosed for simultaneously tracking multiple finger (202-204) and palm (206, 207) contacts as hands approach, touch, and slide across a proximity-sensing, compliant, and flexible multi-touch surface (2). The surface consists of compressible cushion (32), dielectric electrode (33), and circuitry layers. A simple proximity transduction circuit is placed under each electrode to maximize the signal-to-noise ratio and to reduce wiring complexity. Scanning and signal offset removal on electrode array produces low-noise proximity images. Segmentation processing of each proximity image constructs a group of electrodes corresponding to each distinguishable contacts and extracts shape, position and surface proximity features for each group. Groups in successive images which correspond to the same hand contact are linked by a persistent path tracker (245) which also detects individual contact touchdown and liftoff. Classification of intuitive hand configurations and motions enables unprecedented integration of typing, resting, pointing, scrolling, 3D manipulation, and handwriting into a versatile, ergonomic computer input device.

An empirical comparison of supervised learning algorithms

Abstract: A number of supervised learning methods have been introduced in the last decade. Unfortunately, the last comprehensive empirical evaluation of supervised learning was the Statlog Project in the early 90's. We present a large-scale empirical comparison between ten supervised learning methods: SVMs, neural nets, logistic regression, naive bayes, memory-based learning, random forests, decision trees, bagged trees, boosted trees, and boosted stumps. We also examine the effect that calibrating the models via Platt Scaling and Isotonic Regression has on their performance. An important aspect of our study is the use of a variety of performance criteria to evaluate the learning methods.