Yok-Yen Nguwi

Detection and classification of road signs in natural environments

An automatic road sign recognition system first locates road signs within images captured by an imaging sensor on-board of a vehicle, and then identifies the detected road signs. This paper presents an automatic neural-network-based road sign recognition system. First, a study of the existing road sign recognition research is presented. In this study, the issues associated with automatic road sign recognition are described, the existing methods developed to tackle the road sign recognition problem are reviewed, and a comparison of the features of these methods is given. Second, the developed road sign recognition system is described. The system is capable of analysing live colour road scene images, detecting multiple road signs within each image, and classifying the type of road signs detected. The system consists of two modules: detection and classification. The detection module segments the input image in the hue-saturation-intensity colour space, and then detects road signs using a Multi-layer Perceptron neural-network. The classification module determines the type of detected road signs using a series of one to one architectural Multi-layer Perceptron neural networks. Two sets of classifiers are trained using the Resillient-Backpropagation and Scaled-Conjugate-Gradient algorithms. The two modules of the system are evaluated individually first. Then the system is tested as a whole. The experimental results demonstrate that the system is capable of achieving an average recognition hit-rate of 95.96% using the scaled-conjugate-gradient trained classifiers.

Automatic Road Sign Recognition Using Neural Networks

An automatic road sign recognition system first locates road signs within images captured by an imaging sensor on-board of a vehicle, and then identifies road signs assisting the driver of the vehicle to properly operate the vehicle. This paper presents an automatic road sign recognition system capable of analysing live images, detecting multiple road signs within images, and classifying the type of the detected road signs. The system consists of two modules: detection and classification. The detection module segments the input image in the hue-saturation-intensity colour space and locates road signs. The classification module determines the type of detected road signs using a series of one to one architectural Multi Layer Perceptron neural networks. The performances of the classifiers that are trained using Resillient Backpropagation and Scaled Conjugate Gradient algorithms are compared. The experimental results demonstrate that the system is capable of achieving an average recognition hit-rate of 96% using Scaled Conjugate Gradient trained classifiers.

Emergent self-organizing feature map for recognizing road sign images

Road sign recognition system remains a challenging part of designing an Intelligent Driving Support System. While there exist many approaches to classify road signs, none have adopted an unsupervised approach. This paper proposes a way of Self-Organizing feature mapping for recognizing a road sign. The emergent self-organizing map (ESOM) is employed for the feature mapping in this study. It has the capability of visualizing the distance structures as well as the density structure of high-dimensional data sets, in which the ESOM is suitable to detect non-trivial cluster structures. This paper discusses the usage of ESOM for road sign detection and classification. The benchmarking against some other commonly used classifiers was performed. The results demonstrate that the ESOM approach outperforms the others in conducting the same simulations of the road sign recognition. We further demonstrate that the result obtained with ESOM is significantly more superior than traditional SOM which does not take into the boundary effect like ESOM did.

An unsupervised self-organizing learning with support vector ranking for imbalanced datasets

The aim of computational learning algorithm is to establish grounds that work for any types of data, once and for all. However, majority of the classifiers have their base from balanced datasets. This paper discusses the issues related to imbalanced data distribution problem and the common strategy to deal with imbalance datasets. We propose a model capable of handling imbalance datasets well in which other typical classifiers fail to do so. The model adopted a derivation of support vector machines in selecting variables so that the problem of imbalanced data distribution can be relaxed. Then, we used an Emergent Self-Organizing Map (ESOM) to cluster the ranker features so as to provide clusters for unsupervised classification. This work progresses by examining the efficiency of the model in evaluating imbalanced datasets. Experimental results show that the criterion based on weight vector derivative achieves good results and performs consistently well over imbalance datasets. In general, our approach outperforms other classification methods which are unable to handle the imbalanced data distribution in the testing datasets.

Two-tier self-organizing visual model for road sign recognition

This paper attempts to model human brainpsilas cognitive process at the primary visual cortex to comprehend road sign. The cortical maps in visual cortex have been widely focused in recent research. We propose a visual model that locates road sign in an image and identifies the localized road sign. Gabor wavelets are used to encode visual information and extract features. Self-organizing maps are used to cluster and classify the road sign images. We evaluate the system with various test sets. The experimental results show encouraging recognition hit rates. There are quite a number of literatures introducing different approaches to classify road sign, but none has adopted unsupervised approach. This work makes use of two-tier topological maps to recognize road signs. First-tier map, called detecting map, filters out non-road sign images and regions. Second-tier map, called recognizing map, classifies a road sign into appropriate class.

Support vector self-organizing learning for imbalanced medical data

The aim of computational learning algorithm is to establish grounds that works for any types of data, once and for all. However, majority of the classifiers assume the datasets are balanced. This research is targeted towards obtaining a model that is able to handle imbalanced data well. This work progresses by examining the efficiency of the model in evaluating imbalanced medical data. The model adopted a derivation of support vector machines in selecting variables. The classification phase uses unsupervised learning algorithm of Emergent Self-Organizing Map. Experimental results show that the criterion based on weight vector derivative achieves good results and performs consistently well over imbalance data.

Intelligent face locator for smartphone

The market for smartphones is set for serious expansion. Juniper Research [3] is predicting the annual sales of smartphones will swell to 300 million by 2013 - up from about 153 million in year 2008, a rise of around 95 percent. Therefore, the development of mobile phone software well deserves wider attention. This paper introduces a real-time face processing application that was tested in Windows Mobile environment. This work is targeted towards the development of an efficient and intelligent face recognition system. The system is capable of locating the face region using derivative-based filtering, and classifying human face through the use of AdaBoost classifier. The motivation behind this work is that we aim to develop a robust model that can help to locate face for a portable face recognition application. The experiments carried out show that we have achieved the features of mobile application: speed and efficiency, that is able to deploy facial recognition into smartphone.

Number plate recognition in noisy image

Number plate recognition has been used widely for access control, congestion control, vehicle management, security control and vehicle behavior monitoring system. This study discusses the importance of number plate recognition and its corresponding application in different countries. Various methods for recognizing number plates are reviewed. Most of the systems are able to deliver good recognition rate of above 90%. However, there is a lack of literature reporting number plate recognition in images with noisy background. We propose and report a system that is able to tolerate noise level up to 20% with recognition rate of 85%. The system utilized a combination of filters and morphological transformation for segmenting the number plate. It then uses resilient back-propagation neural networks for recognition.

Support-vector-based emergent self-organising approach for emotional understanding

This study discusses the computational analysis of general emotion understanding from questionnaires methodology. The questionnaires method approaches the subject by investigating the real experience that accompanied the emotions, whereas the other laboratory approaches are generally associated with exaggerated elements. We adopted a connectionist model called support-vector-based emergent self-organising map (SVESOM) to analyse the emotion profiling from the questionnaires method. The SVESOM first identifies the important variables by giving discriminative features with high ranking. The classifier then performs the classification based on the selected features. Experimental results show that the top rank features are in line with the work of Scherer and Wallbott [(1994), ‘Evidence for Universality and Cultural Variation of Differential Emotion Response Patterning’, Journal of Personality and Social Psychology, 66, 310–328], which approached the emotions physiologically. While the performance measures show that using the full features for classifications can degrade the performance, the selected features provide superior results in terms of accuracy and generalisation.

Towards a portable intelligent facial expression recognizer

Facial expression recognition is a challenging task. A facial expression is fonned by contracting or relaxing different facial muscles on human face which results in temporally deformed facial features like wide open mouth, raising eyebrows or etc. Such a system presents challenges. For instances, lighting condition is a very difficult problem to constraint and regulate. On the other hand, real-time processing is also a challenging problem since there are so many facial features to be extracted and processed and sometime conventional classifiers are not even effective to handle those features and then produce good classification perfonnance. This paper discusses the issues on how the advanced feature selection techniques together with good classifiers can playa vital important role of real-time facial expression recognition. The content of this paper is a way to open-up a discussion about building a real-time system to read and respond to the emotions of people from facial expressions.