Spectral–texture feature extraction using statistical moments with application to object-based vegetation species classification

The use of appropriate features to characterise an output class or object is critical for all classification problems. In order to find optimal feature descriptors for vegetation species classification in a power line corridor monitoring application, this article evaluates the capability of several spectral and texture features. A new idea of spectral–texture feature descriptor is proposed by incorporating spectral vegetation indices in statistical moment features. The proposed method is evaluated against several classic texture feature descriptors. Object-based classification method is used and a support vector machine is employed as the benchmark classifier. Individual tree crowns are first detected and segmented from aerial images and different feature vectors are extracted to represent each tree crown. The experimental results showed that the proposed spectral moment features outperform or can at least compare with the state-of-the-art texture descriptors in terms of classification accuracy. A comprehensive quantitative evaluation using receiver operating characteristic space analysis further demonstrates the strength of the proposed feature descriptors.

Multi-modal summarization of key events and top players in sports tournament videos

To detect and annotate the key events of live sports videos, we need to tackle the semantic gaps of audio-visual information. Previous work has successfully extracted semantic from the time-stamped web match reports, which are synchronized with the video contents. However, web and social media articles with no time-stamps have not been fully leveraged, despite they are increasingly used to complement the coverage of major sporting tournaments. This paper aims to address this limitation using a novel multimodal summarization framework that is based on sentiment analysis and players' popularity. It uses audiovisual contents, web articles, blogs, and commentators' speech to automatically annotate and visualize the key events and key players in a sports tournament coverage. The experimental results demonstrate that the automatically generated video summaries are aligned with the events identified from the official website match reports.

Digital signal processing

Signal Processing (SP) is a subject of central importance in engineering and the applied sciences. Signals are information-bearing functions, and SP deals with the analysis and processing of signals (by dedicated systems) to extract or modify information. Signal processing is necessary because signals normally contain information that is not readily usable or understandable, or which might be disturbed by unwanted sources such as noise. Although many signals are non-electrical, it is common to convert them into electrical signals for processing. Most natural signals (such as acoustic and biomedical signals) are continuous functions of time, with these signals being referred to as analog signals. Prior to the onset of digital computers, Analog Signal Processing (ASP) and analog systems were the only tool to deal with analog signals. Although ASP and analog systems are still widely used, Digital Signal Processing (DSP) and digital systems are attracting more attention, due in large part to the significant advantages of digital systems over the analog counterparts. These advantages include superiority in performance,s peed, reliability, efficiency of storage, size and cost. In addition, DSP can solve problems that cannot be solved using ASP, like the spectral analysis of multicomonent signals, adaptive filtering, and operations at very low frequencies. Following the recent developments in engineering which occurred in the 1980's and 1990's, DSP became one of the world's fastest growing industries. Since that time DSP has not only impacted on traditional areas of electrical engineering, but has had far reaching effects on other domains that deal with information such as economics, meteorology, seismology, bioengineering, oceanology, communications, astronomy, radar engineering, control engineering and various other applications. This book is based on the Lecture Notes of Associate Professor Zahir M. Hussain at RMIT University (Melbourne, 2001-2009), the research of Dr. Amin Z. Sadik (at QUT & RMIT, 2005-2008), and the Note of Professor Peter O'Shea at Queensland University of Technology. Part I of the book addresses the representation of analog and digital signals and systems in the time domain and in the frequency domain. The core topics covered are convolution, transforms (Fourier, Laplace, Z. Discrete-time Fourier, and Discrete Fourier), filters, and random signal analysis. There is also a treatment of some important applications of DSP, including signal detection in noise, radar range estimation, banking and financial applications, and audio effects production. Design and implementation of digital systems (such as integrators, differentiators, resonators and oscillators are also considered, along with the design of conventional digital filters. Part I is suitable for an elementary course in DSP. Part II (which is suitable for an advanced signal processing course), considers selected signal processing systems and techniques. Core topics covered are the Hilbert transformer, binary signal transmission, phase-locked loops, sigma-delta modulation, noise shaping, quantization, adaptive filters, and non-stationary signal analysis. Part III presents some selected advanced DSP topics. We hope that this book will contribute to the advancement of engineering education and that it will serve as a general reference book on digital signal processing.

Designing for engagement towards healthier lifestyles through food image sharing : the case of I8DAT

This paper introduces the underlying design concepts of I8DAT, a food image sharing application that has been developed as part of a three-year research project – Eat, Cook, Grow: Ubiquitous Technology for Sustainable Food Culture in the City (http://www.urbaninformatics .net/projects/food) – exploring urban food practices to engage people in healthier, more environmentally and socially sustainable eating, cooking, and growing food in their everyday lives. The key aim of the project is to produce actionable knowledge, which is then applied to create and test several accessible, user-centred interactive design solutions that motivate user-engagement through playful and social means rather than authoritative information distribution. Through the design and implementation processes we envisage to integrate these design interventions to create a sustainable food network that is both technical and socio-cultural in nature (technosocial). Our primary research locale is Brisbane, Australia, with additional work carried out in three reference cities with divergent geographic, socio-cultural, and technological backgrounds: Seoul, South Korea, for its global leadership in ubiquitous technology, broadband access, and high population density; Lincoln, UK, for the regional and peri-urban dimension it provides, and Portland, Oregon, US, for its international standing as a hub of the sustainable food movement.

Facial expression recognition using facial movement features

Facial expression is an important channel for human communication and can be applied in many real applications. One critical step for facial expression recognition (FER) is to accurately extract emotional features. Current approaches on FER in static images have not fully considered and utilized the features of facial element and muscle movements, which represent static and dynamic, as well as geometric and appearance characteristics of facial expressions. This paper proposes an approach to solve this limitation using ‘salient’ distance features, which are obtained by extracting patch-based 3D Gabor features, selecting the ‘salient’ patches, and performing patch matching operations. The experimental results demonstrate high correct recognition rate (CRR), significant performance improvements due to the consideration of facial element and muscle movements, promising results under face registration errors, and fast processing time. The comparison with the state-of-the-art performance confirms that the proposed approach achieves the highest CRR on the JAFFE database and is among the top performers on the Cohn-Kanade (CK) database.

Improving the performance of facial expression recognition using dynamic, subtle and regional features

Human facial expression is a complex process characterized of dynamic, subtle and regional emotional features. State-of-the-art approaches on facial expression recognition (FER) have not fully utilized this kind of features to improve the recognition performance. This paper proposes an approach to overcome this limitation using patch-based ‘salient’ Gabor features. A set of 3D patches are extracted to represent the subtle and regional features, and then inputted into patch matching operations for capturing the dynamic features. Experimental results show a significant performance improvement of the proposed approach due to the use of the dynamic features. Performance comparison with pervious work also confirms that the proposed approach achieves the highest CRR reported to date on the JAFFE database and a top-level performance on the Cohn-Kanade (CK) database.

The sample complexity of pattern classification with neural networks: the size of the weights is more important than the size of the network

Sample complexity results from computational learning theory, when applied to neural network learning for pattern classification problems, suggest that for good generalization performance the number of training examples should grow at least linearly with the number of adjustable parameters in the network. Results in this paper show that if a large neural network is used for a pattern classification problem and the learning algorithm finds a network with small weights that has small squared error on the training patterns, then the generalization performance depends on the size of the weights rather than the number of weights. For example, consider a two-layer feedforward network of sigmoid units, in which the sum of the magnitudes of the weights associated with each unit is bounded by A and the input dimension is n. We show that the misclassification probability is no more than a certain error estimate (that is related to squared error on the training set) plus A3 √((log n)/m) (ignoring log A and log m factors), where m is the number of training patterns. This may explain the generalization performance of neural networks, particularly when the number of training examples is considerably smaller than the number of weights. It also supports heuristics (such as weight decay and early stopping) that attempt to keep the weights small during training. The proof techniques appear to be useful for the analysis of other pattern classifiers: when the input domain is a totally bounded metric space, we use the same approach to give upper bounds on misclassification probability for classifiers with decision boundaries that are far from the training examples.

Model selection and error estimation

We study model selection strategies based on penalized empirical loss minimization. We point out a tight relationship between error estimation and data-based complexity penalization: any good error estimate may be converted into a data-based penalty function and the performance of the estimate is governed by the quality of the error estimate. We consider several penalty functions, involving error estimates on independent test data, empirical VC dimension, empirical VC entropy, and margin-based quantities. We also consider the maximal difference between the error on the first half of the training data and the second half, and the expected maximal discrepancy, a closely related capacity estimate that can be calculated by Monte Carlo integration. Maximal discrepancy penalty functions are appealing for pattern classification problems, since their computation is equivalent to empirical risk minimization over the training data with some labels flipped.

Infinite-horizon policy-gradient estimation

Gradient-based approaches to direct policy search in reinforcement learning have received much recent attention as a means to solve problems of partial observability and to avoid some of the problems associated with policy degradation in value-function methods. In this paper we introduce GPOMDP, a simulation-based algorithm for generating a biased estimate of the gradient of the average reward in Partially Observable Markov Decision Processes (POMDPs) controlled by parameterized stochastic policies. A similar algorithm was proposed by Kimura, Yamamura, and Kobayashi (1995). The algorithm's chief advantages are that it requires storage of only twice the number of policy parameters, uses one free parameter β ∈ [0,1) (which has a natural interpretation in terms of bias-variance trade-off), and requires no knowledge of the underlying state. We prove convergence of GPOMDP, and show how the correct choice of the parameter β is related to the mixing time of the controlled POMDP. We briefly describe extensions of GPOMDP to controlled Markov chains, continuous state, observation and control spaces, multiple-agents, higher-order derivatives, and a version for training stochastic policies with internal states. In a companion paper (Baxter, Bartlett, & Weaver, 2001) we show how the gradient estimates generated by GPOMDP can be used in both a traditional stochastic gradient algorithm and a conjugate-gradient procedure to find local optima of the average reward. ©2001 AI Access Foundation and Morgan Kaufmann Publishers. All rights reserved.

Learning the kernel matrix with semidefinite programming

Kernel-based learning algorithms work by embedding the data into a Euclidean space, and then searching for linear relations among the embedded data points. The embedding is performed implicitly, by specifying the inner products between each pair of points in the embedding space. This information is contained in the so-called kernel matrix, a symmetric and positive semidefinite matrix that encodes the relative positions of all points. Specifying this matrix amounts to specifying the geometry of the embedding space and inducing a notion of similarity in the input space - classical model selection problems in machine learning. In this paper we show how the kernel matrix can be learned from data via semidefinite programming (SDP) techniques. When applied to a kernel matrix associated with both training and test data this gives a powerful transductive algorithm -using the labeled part of the data one can learn an embedding also for the unlabeled part. The similarity between test points is inferred from training points and their labels. Importantly, these learning problems are convex, so we obtain a method for learning both the model class and the function without local minima. Furthermore, this approach leads directly to a convex method for learning the 2-norm soft margin parameter in support vector machines, solving an important open problem.

Rademacher and Gaussian complexities: Risk bounds and structural results

We investigate the use of certain data-dependent estimates of the complexity of a function class, called Rademacher and Gaussian complexities. In a decision theoretic setting, we prove general risk bounds in terms of these complexities. We consider function classes that can be expressed as combinations of functions from basis classes and show how the Rademacher and Gaussian complexities of such a function class can be bounded in terms of the complexity of the basis classes. We give examples of the application of these techniques in finding data-dependent risk bounds for decision trees, neural networks and support vector machines.