Associate Professor Suhuai Luo

The proliferation of fake news on social media platforms poses significant challenges to society and individuals, leading to negative impacts. As the tactics employed by purveyors of fake news continue to evolve, there is an urgent need for automatic fake news detection (FND) to mitigate its adverse social consequences. Machine learning (ML) and deep learning (DL) techniques have emerged as promising approaches for characterising and identifying fake news content. This paper presents an extensive review of previous studies aiming to understand and combat the dissemination of fake news. The review begins by exploring the definitions of fake news proposed in the literature and delves into related terms and psychological and scientific theories that shed light on why people believe and disseminate fake news. Subsequently, advanced ML and DL techniques for FND are dicussed in detail, focusing on three main feature categories: content-based, context-based, and hybrid-based features. Additionally, the review summarises the characteristics of fake news, commonly used datasets, and the methodologies employed in existing studies. Furthermore, the review identifies the challenges current FND studies encounter and highlights areas that require further investigation in future research. By offering a comprehensive overview of the field, this survey aims to serve as a guide for researchers working on FND, providing valuable insights for developing effective FND mechanisms in the era of technological advancements.

Malware detection approaches can be classified into two classes, including static analysis and dynamic analysis. Conventional approaches of the two classes have their respective advantages and disadvantages. For example, static analysis is faster but cannot detect the malware variants generated through code obfuscation, whereas dynamic analysis can effectively detect variants generated through code obfuscation but is slower and requires intensive resources. This paper proposes a novel deep learning-based approach for malware detection. It delivers better performance than conventional approaches by combining static and dynamic analysis advantages. First, it visualises a portable executable (PE) file as a coloured image. Second, it extracts deep features from the colour image using fine-tuned deep learning model. Third, it detects malware based on the deep features using support vector machines (SVM). The proposed method combines deep learning with machine learning and eliminates the need for intensive feature engineering tasks and domain knowledge. The proposed approach is scalable, cost-effective, and efficient. The detection effectiveness of the proposed method is validated through 12 machine learning models and 15 deep learning models. The generalisability of the proposed framework is validated on various benchmark datasets. The proposed approach outperformed with an accuracy of 99.06% on the Malimg dataset. The Wilcoxon signed-rank test is used to show the statistical significance of the proposed framework. The detailed experimental results demonstrate the superiority of the proposed method over the other state-of-the-art approaches, with an average increase in accuracy of 16.56%. Finally, to tackle the problems of imbalanced data and the shortage of publicly available datasets for malware detection, various data augmentation techniques are proposed, which lead to improved performance. It is evident from the results that the proposed framework can be useful to the defence industry, which will be helpful in devising more efficient malware detection solutions.

Surveillance videos are crucial in imparting public security, reducing or avoiding the accidents that occur from anomalies. Crowd anomaly detection is a rapidly growing research field that aims to identify abnormal or suspicious behavior in crowds. This paper provides a comprehensive review of the state-of-the-art in crowd anomaly detection and, different taxonomies, publicly available datasets, challenges, and future research directions. The paper first provides an overview of the field and the importance of crowd anomaly detection in various applications such as public safety, transportation, and surveillance. Secondly, it presents the components of crowd anomaly detection and its different taxonomies based on the availability of labels, and the type of anomalies. Thirdly, it presents the review of the recent progress of crowd anomaly detection. The review also covers publicly available datasets commonly used for evaluating crowd anomaly detection methods. The challenges faced by the field, such as handling variability in crowd behavior, dealing with large and complex data sets, and addressing the imbalance of data, are discussed. Finally, the paper concludes with a discussion of future research directions in crowd anomaly detection, including integrating multiple modalities, addressing privacy concerns, and addressing crowd monitoring systems¿ ethical and legal implications.

Skin cancer can be fatal if it is found to be malignant. Modern diagnosis of skin cancer heavily relies on visual inspection through clinical screening, dermoscopy, or histopathological examinations. However, due to similarity among cancer types, it is usually challenging to identify the type of skin cancer, especially at its early stages. Deep learning techniques have been developed over the last few years and have achieved success in helping to improve the accuracy of diagnosis and classification. However, the latest deep learning algorithms still do not provide ideal classification accuracy. To further improve the performance of classification accuracy, this paper presents a novel method of classifying skin cancer in clinical skin images. The method consists of four blocks. First, class rebalancing is applied to the images of seven skin cancer types for better classification performance. Second, an image is preprocessed by being split into patches of the same size and then flattened into a series of tokens. Third, a transformer encoder is used to process the flattened patches. The transformer encoder consists of N identical layers with each layer containing two sublayers. Sublayer one is a multihead self-attention unit, and sublayer two is a fully connected feed-forward network unit. For each of the two sublayers, a normalization operation is applied to its input, and a residual connection of its input and its output is calculated. Finally, a classification block is implemented after the transformer encoder. The block consists of a flattened layer and a dense layer with batch normalization. Transfer learning is implemented to build the whole network, where the ImageNet dataset is used to pretrain the network and the HAM10000 dataset is used to fine-tune the network. Experiments have shown that the method has achieved a classification accuracy of 94.1%, outperforming the current state-of-the-art model IRv2 with soft attention on the same training and testing datasets. On the Edinburgh DERMOFIT dataset also, the method has better performance compared with baseline models.

Efforts have been dedicated by researchers in the field of natural language processing (NLP) to detecting and combating fake news using an assortment of machine learning (ML) and deep learning (DL) techniques. In this paper, a review of the existing studies is conducted to understand and curtail the dissemination of fake news. Specifically, we conducted a benchmark study using a wide range of (1) classical ML algorithms such as logistic regression (LR), support vector machines (SVM), decision tree (DT), naive Bayes (NB), random forest (RF), XGBoost (XGB) and an ensemble learning method of such algorithms, (2) advanced ML algorithms such as convolutional neural networks (CNNs), bidirectional long short-term memory (BiLSTM), bidirectional gated recurrent units (BiGRU), CNN-BiLSTM, CNN-BiGRU and a hybrid approach of such techniques and (3) DL transformer-based models such as BERT (Formula presented.) and RoBERTa (Formula presented.). The experiments are carried out using different pretrained word embedding methods across four well-known real-world fake news datasets¿LIAR, PolitiFact, GossipCop and COVID-19¿to examine the performance of different techniques across various datasets. Furthermore, a comparison is made between context-independent embedding methods (e.g., GloVe) and the effectiveness of BERT (Formula presented.) ¿contextualised representations in detecting fake news. Compared with the state of the art¿s results across the used datasets, we achieve better results by solely relying on news text. We hope this study can provide useful insights for researchers working on fake news detection.

Disease diagnosis is a challenging task due to a large number of associated factors. Uncertainty in the diagnosis process arises from inaccuracy in patient attributes, missing data, and limitation in the medical expert¿s ability to define cause and effect relationships when there are multiple interrelated variables. This paper aims to demonstrate an integrated view of deploying smart disease diagnosis using the Internet of Things (IoT) empowered by the fuzzy inference system (FIS) to diagnose various diseases. The Fuzzy System is one of the best systems to diagnose medical conditions because every disease diagnosis involves many uncertainties, and fuzzy logic is the best way to handle uncertainties. Our proposed system differentiates new cases provided symptoms of the disease. Generally, it becomes a time-sensitive task to discriminate symptomatic diseases. The proposed system can track symptoms firmly to diagnose diseases through IoT and FIS smartly and efficiently. Different coefficients have been employed to predict and compute the identified disease¿s severity for each sign of disease. This study aims to differentiate and diagnose COVID-19, Typhoid, Malaria, and Pneumonia. This study used the FIS method to figure out the disease over the use of given data related to correlating with input symptoms. MATLAB tool is utilised for the implementation of FIS. Fuzzy procedure on the aforementioned given data presents that affectionate disease can derive from the symptoms. The results of our proposed method proved that FIS could be utilised for the diagnosis of other diseases. This study may assist doctors, patients, medical practitioners, and other healthcare professionals in early diagnosis and better treat diseases.

Pneumoconiosis is a group of occupational lung diseases induced by mineral dust inhalation and subsequent lung tissue reactions. It can eventually cause irreparable lung damage, as well as gradual and permanent physical impairments. It has affected millions of workers in hazardous industries throughout the world, and it is a leading cause of occupational death. It is difficult to diagnose early pneumoconiosis because of the low sensitivity of chest radiographs, the wide variation in interpretation between and among readers, and the scarcity of B-readers, which all add to the difficulty in diagnosing these occupational illnesses. In recent years, deep machine learning algorithms have been extremely successful at classifying and localising abnormality of medical images. In this study, we proposed an ensemble learning approach to improve pneumoconiosis detection in chest X-rays (CXRs) using nine machine learning classifiers and multi-dimensional deep features extracted using CheXNet-121 architecture. There were eight evaluation metrics utilised for each high-level feature set of the associated cross-validation datasets in order to compare the ensemble performance and state-of-the-art techniques from the literature that used the same cross-validation datasets. It is observed that integrated ensemble learning exhibits promising results (92.68% accuracy, 85.66% Matthews correlation coefficient (MCC), and 0.9302 area under the precision¿recall (PR) curve), compared to individual CheXNet-121 and other state-of-the-art techniques. Finally, Grad-CAM was used to visualise the learned behaviour of individual dense blocks within CheXNet-121 and their ensembles into three-color channels of CXRs. We compared the Grad-CAM-indicated ROI to the ground-truth ROI using the intersection of the union (IOU) and average-precision (AP) values for each classifier and their ensemble. Through the visualisation of the Grad-CAM within the blue channel, the average IOU passed more than 90% of the pneumoconiosis detection in chest radiographs.

Multiple sclerosis (MS) is a chronic neurological disease of the central nervous system (CNS). Early diagnosis of MS is highly desirable as treatments are more effective in preventing MS-related disability when given in the early stages of the disease. The main aim of this research is to predict the occurrence of a second MS-related clinical event, which indicates the conversion of clinically isolated syndrome (CIS) to clinically definite MS (CDMS). In this study, we apply a branch of artificial intelligence known as deep learning and develop a fully automated algorithm primed with convolutional neural network (CNN) that has the ability to learn from MRI scan features. The basic architecture of our algorithm is that of the VGG16 CNN model, but amended such that it can handle MRI DICOM images. A dataset comprised of scans acquired using two different scanners was used for the purposes of verification of the algorithm. A group of 49 patients had volumetric MRI scans taken at onset of the disease and then again one year later using one of the two scanners. In total, this yielded 7360 images which were then used for training, validation, and testing of the algorithm. Initially, these raw images were taken through 4 steps of preprocessing. In order to boost the efficiency of the process, we pretrained our algorithm using the publicly available ADNI dataset used to classify Alzheimer's disease. Finally, we used our preprocessed dataset to train and test the algorithm. Clinical evaluation conducted a year after the first time point revealed that 26 of the 49 patients had converted to CDMS, while the remaining 23 had not. Results of testing showed that our algorithm was able to predict the clinical results with an accuracy of 88.8% and with an area under the curve (AUC) of 91%. A highly accurate algorithm was developed using CNN approach to reliably predict conversion of patients with CIS to CDMS using MRI data from two different scanners.

Background: Computer-aided diagnosis can facilitate the early detection and diagnosis of multiple sclerosis (MS) thus enabling earlier interventions and a reduction in long-term MS-related disability. Recent advancements in the field of artificial intelligence (AI) have led to the improvements in the classification, quantification and identification of diagnostic patterns in medical images for a range of diseases, in particular, for MS. Importantly, data generated using AI techniques are analyzed automatically, which compares favourably with labour-intensive and time-consuming manual methods. Objective: The aim of this review is to assist MS researchers to understand current and future developments in the AI-based diagnosis and prognosis of MS. Methods: We will investigate a variety of AI approaches and various classifiers and compare the current state-of-the-art techniques in relation to lesion segmentation/detection and prognosis of disease. After briefly describing the magnetic resonance imaging (MRI) techniques commonly used, we will describe AI techniques used for the detection of lesions and MS prognosis. Results: We then evaluate the clinical maturity of these AI techniques in relation to MS. Conclusion: Finally, future research challenges are identified in a bid to encourage further improvements of the methods.

Early detection of pneumoconiosis in X-Rays has been a challenging task that leads to high inter- and intra-reader variability. Motivated by the success of deep learning in general and medical image classification, this paper proposes an approach to automatically detect pneumoconiosis using a deep feature based binary classifier. The features are extracted from X-rays using deep transfer learning, comprising both low and high-level feature sets. For this, a CNN model pre-trained with a transfer learning from a CheXNet model was initially used to extract deep features from the X-Ray images, then the deep features were mapped to higher-dimensional feature spaces for classification using Support Vector Machine (SVM) and CNN based feature aggregation methods. In order to cross validate the proposed method, the training and testing images were randomly split into three folds before each experiment. Nine evaluation metrics were employed to compare the performance of the proposed method and state-of-the-art methods from the literature that used the same datasets. The experimental results show that the proposed framework outperformed others, achieving an accuracy of 92.68% in the automated detection of pneumoconiosis.

Background: Alzheimer's disease (AD) is one of the deadliest diseases in developed countries. Treatments following early AD detection can significantly delay institutionalisation and extend patients' independence. There has been a growing focus on early AD detection using artificial intelligence. Convolutional neural networks (CNNs) have proven revolutionary for image-based applications and have been applied to brain scans. In recent years, studies have utilised two-dimensional (2D) CNNs on magnetic resonance imaging (MRI) scans for AD detection. To apply a 2D CNN on three-dimensional (3D) MRI volumes, each MRI scan is split into 2D image slices. A CNN is trained over the image slices by calculating a loss function between each subject's label and each image slice's predicted output. Although 2D CNNs can discover spatial dependencies in an image slice, they cannot understand the temporal dependencies among 2D image slices in a 3D MRI volume. This study aims to resolve this issue by modelling the sequence of MRI features produced by a CNN with deep sequence-based networks for AD detection. Method: The CNN utilised in this paper was ResNet-18 pre-trained on an ImageNet dataset. The employed sequence-based models were the temporal convolutional network (TCN) and different types of recurrent neural networks. Several deep sequence-based models and configurations were implemented and compared for AD detection. Results: Our proposed TCN model achieved the best classification performance with 91.78% accuracy, 91.56% sensitivity and 92% specificity. Conclusion: Our results show that applying sequence-based models can improve the classification accuracy of 2D and 3D CNNs for AD detection by up to 10%.

In our work, we have presented two widely used recommendation systems. We have presented a context-aware recommender system to filter the items associated with user¿s interests coupled with a context-based recommender system to prescribe those items. In this study, context-aware recommender systems perceive the user¿s location, time, and company. The context-based recommender system retrieves patterns from World Wide Web-based on the user¿s past interactions and provides future news recommendations. We have presented different techniques to support media recommendations for smartphones, to create a framework for context-aware, to filter E-learning content, and to deliver convenient news to the user. To achieve this goal, we have used content-based, collaborative filtering, a hybrid recommender system, and implemented a Web ontology language (OWL). We have also used the Resource Description Framework (RDF), JAVA, machine learning, semantic mapping rules, and natural ontology languages that suggest user items related to the search. In our work, we have used E-paper to provide users with the required news. After applying the semantic reasoning approach, we have concluded that by some means, this approach works similarly as a content-based recommender system since by taking the gain of a semantic approach, we can also recommend items according to the user¿s interests. In a content-based recommender system, the system provides additional options or results that rely on the user¿s ratings, appraisals, and interests.

Malignant mesothelioma (MM) is a rare cancer type arising from mesothelial cells. The current clinical diagnosis is based on contrast-enhanced computed tomography, magnetic resonance imaging, and positron emission tomography that are either invasive or costly. The failure to diagnose malignantly can lead to an increased risk of multiple medical conditions, including cardiovascular diseases, emotional distress, anemia, and diabetes. To date, there is a limited number of prognostic factors that can be used for diagnosis. Most existing work has considered the MM disease as a classification task. In contrast, our study has initiated a knowledge extraction problem and proposed a machine learning-based framework. The performance status, age, and sex of patients are currently the most substantial clinical prognostic factors, but other histopathological and clinical prognostic factors are still unclear. This study aims to search for clinical prognostic, radiological, and histopathological factors in MM. In this study, the latest dataset from a public repository (UCI) has been utilised, including patients' medical, socio-economic, histopathological, and clinical factors. Association rule mining-based algorithms (Apriori and frequent pattern (FP) growth method) and feature selection techniques have been employed to extract significant features. The performance of the proposed framework has been evaluated based on support, confidence, and lift. We set the support, confidence, and lift between 0.5¿1.0, 0.5¿1.0, and 1.0¿1.6 respectively. Our results showed five significant prognosis factors with the values for the identification of MM: Pleural lactate dehydrogenase >500 IU/L, C-reactive protein >10/µL, pleural albumin<3/µL, the presence of asbestos exposure and pleural effusion. In nearly all the experiments, the binary features were among the leading top five features in the list. The diagnosis of MM can be accessible through prognostic factors. Our proposed framework will help to diagnose the patients without expensive tests and painful procedures. The proposed framework may assist doctors, patients, medical practitioners, and other healthcare professionals for early diagnosis and better treatment of malignant mesothelioma through significant prognostic factors.

The contemporary innovations in financial technology (fintech) serve society with an environmentally friendly atmosphere. Fintech covers an enormous range of activities from data security to financial service deliverables that enable the companies to automate their existing business structure and introduce innovative products and services. Therefore, there is an increasing demand for scholars and professionals to identify the future trends and directions of the topic. This is why the present study conducted a bibliometric analysis in social, environmental, and computer sciences fields to analyse the implementation of environment-friendly computer applications to benefit societal growth and well-being. We have used the ¿bibliometrix 3.0¿ package of the r-program to analyse the core aspects of fintech systematically. The study suggests that ¿ACM International Conference Proceedings¿ is the core source of published fintech literature. China leads in both multiple and single country production of fintech publications. Bina Nusantara University is the most relevant affiliation. Arner and Buckley provide impactful fintech literature. In the conceptual framework, we analyse relationships between different topics of fintech and address dynamic research streams and themes. These research streams and themes highlight the future directions and core topics of fintech. The study deploys a co-occurrence network to differentiate the entire fintech literature into three research streams. These research streams are related to ¿cryptocurrencies, smart contracts, financial technology¿, ¿financial industry stability, service, innovation, regulatory technology (regtech)¿, and ¿machine learning and deep learning innovations¿. The study deploys a thematic map to identify basic, emerging, dropping, isolated, and motor themes based on centrality and density. These various themes and streams are designed to lead the researchers, academicians, policymakers, and practitioners to narrow, distinctive, and significant topics.

The year 2020 has changed the living style of people all around the world. Corona pandemic has affected the people in all fields of life economically, physically, and mentally. This dataset is a collection of published articles discussing the effect of COVID and SARS on the social sciences from 2003 to 2020. This dataset collection and analysis highlight the significance and influential aspects, research streams, and themes in this domain. The analysis provides top journals, highly cited articles, mostly used keywords, top affiliation institutes, leading countries based on the citation, potential research streams, a thematic map, and future directions in this area of research. In the future, this dataset will be helpful for every researcher and policymakers to proceed as a starting point to identify the relevant research based on the analysis of 18 years of research in this domain.

General radiographs are an initial diagnostic tool for a variety of clinical conditions such as lung disease detection. The size, shape and texture of a lung field are key parameters for X-ray radiographs based lung disease diagnosis in which the lung field segmentation is a significant step. Although many new methods have been proposed in medical image applications, the lung field segmentation remains a challenge. In this paper, we have proposed an improved segmentation method based on statistical shape and appearance models. For the shape model, multi-scale and multi-step-size with different limitation parameters were used to increase the searching ability. For the appearance model, multiple features with different weights were used to describe different parts of the lung field border. A set of 247 chest radiographs was used to test the method. The average overlap of the proposed method was 93.1% for the publicly available JSRT database. The experiment results show that the proposed method outperforms other active shape model based methods.

In this paper, a customized active shape model to extract lungs from radiography chest images was proposed and validated. Firstly, the average active shape model, gray-scale projection and affine registration were employed to attain the initial lung contours. Secondly, a new objective function with constraints of distance and edge was proposed to push the vertices of active shape model to the real lung edge, pull the vertices out of the stomach gas regions, and have a more balanced distance distribution of vertices. Finally, multi-resolution representation and optimization were employed to attain fast optimization. Experimental results on a public database of 247 images showed that the proposed algorithm could achieve an average accuracy of 94.7%, which is 4.4% better than the traditional active shape model and 2.7% better than the active shape model with local invariant features.

Emotional factors directly reflect audiences' attention, evaluation and memory. Recently, video affective content analysis attracts more and more research efforts. Most of the existing methods map low-level affective features directly to emotions by applying machine learning. Compared to human perception process, there is actually a gap between low-level features and high-level human perception of emotion. In order to bridge the gap, we propose a three-level affective content analysis framework by introducing mid-level representation to indicate dialog, audio emotional events (e.g., horror sounds and laughters) and textual concepts (e.g., informative keywords). Mid-level representation is obtained from machine learning on low-level features and used to infer high-level affective content. We further apply the proposed framework and focus on a number of case studies. Audio emotional event, dialog and subtitle are studied to assist affective content detection in different video domains/genres. Multiple modalities are considered for affective analysis, since different modality has its own merit to evoke emotions. Experimental results shows the proposed framework is effective and efficient for affective content analysis. Audio emotional event, dialog and subtitle are promising mid-level representations. © 2012 Springer Science+Business Media, LLC.

The management of electricity system in home environments plays an important role in generating energy consumption and improving efficiency of energy usage. At present, nonintrusive appliance load monitoring (NIALM) techniques are the most effective approach for estimating the electrical power consumption of individual appliances. This paper presents our contribution in intelligent electrical appliance decomposition in home environment. It is a modified power appliance disaggregation technique based on power harmonic features and support vector machine (SVM). It has higher recognition accuracy and faster computational speed. The experimental results of the power decomposition technique on real date are presented with promising results. © (2013) Trans Tech Publications, Switzerland.

We present a cerebral vasculature extraction method from magnetic resonance angiography (MRA) and provide validation for arteries. After reviewing the state-of-the-art in vasculature segmentation techniques, we introduce an automatic algorithm with robust maximal intensity searching and region growing. We present the details of the design and application of extraction validation interface. We demonstrate the artery extraction fidelity of the method with tests on both 3D phantom and real MRA images. We conclude by summarising the proposed algorithm and pointing out possible future pursuits in vessel segmentation. © 2005.

The present-day world has become all dependent on cyberspace for every aspect of daily living. The use of cyberspace is rising with each passing day. The world is spending more time on the Internet than ever before. As a result, the risks of cyber threats and cybercrimes are increasing. The term 'cyber threat' is referred to as the illegal activity performed using the Internet. Cybercriminals are changing their techniques with time to pass through the wall of protection. Conventional techniques are not capable of detecting zero-day attacks and sophisticated attacks. Thus far, heaps of machine learning techniques have been developed to detect the cybercrimes and battle against cyber threats. The objective of this research work is to present the evaluation of some of the widely used machine learning techniques used to detect some of the most threatening cyber threats to the cyberspace. Three primary machine learning techniques are mainly investigated, including deep belief network, decision tree and support vector machine. We have presented a brief exploration to gauge the performance of these machine learning techniques in the spam detection, intrusion detection and malware detection based on frequently used and benchmark datasets.

We live in a world where data is expanding exponentially. Most of the data is unstructured when obtained through the web. Many organizations, institutes, and governments worldwide gather public views regarding their products, services, or policies. With thousands of reviews about some product, service, or policy, it is impossible to conclude some kind of final thought from it. To handle this, there is a desperate need for a model that can extract meaningful information from data to make correct and timely decisions for the efficient growth of business and smooth running of an organization or government. Otherwise, the practice of collecting and storing data will be ineffective. In this study, we focused on conducting an extensive public survey on issues of Southern Punjab, carry out appropriate processing on collected data and predict trends in public opinion for decision-making. Natural Language Processing (NLP) and Machine Learning (ML) have dealt with this problem. Different data preprocessing techniques have been utilized to remove the noise from data. Our experiments stated that unemployment, poverty, education, and corruption are the major issues of the targeted region. This study will help government officials and non-governmental organizations to be focused on the extracted issues in the specific region.

This paper focuses on detecting Alzheimer's Disease (AD) using the ResNet-18 model on Magnetic Resonance Imaging (MRI). Previous studies have applied different 2D Convolutional Neural Networks (CNNs) to detect AD. The main idea being to split 3D MRI scans into 2D image slices, so that classification can be performed on the image slices independently. This idea allows researchers to benefit from the concept of transfer learning. However, 2D CNNs are incapable of understanding the relationship among 2D image slices in a 3D MRI scan. One solution is to employ 3D CNNs instead of 2D ones. In this paper, we propose a method to utilise transfer learning in 3D CNNs, which allows the transfer of knowledge from 2D image datasets to a 3D image dataset. Both 2D and 3D CNNs are compared in this study, and our results show that introducing transfer learning to a 3D CNN improves the accuracy of an AD detection system. After using an optimisation method in the training process, our approach achieved 96.88% accuracy, 100% sensitivity, and 93.75% specificity.

The management of electricity system in home environments plays an important role in managing energy consumption and improving efficiency of energy usage. At present, nonintrusive appliance load monitoring (NIALM) techniques are the most effective approach for estimating the electrical power consumption of individual appliances. This paper presents our contributions in intelligent electrical appliance recognition in home environment. The novel method is for general power load classification and disaggregation which is mainly carried out by combining support vector machines with various power features. The experiments on real world data have demonstrated higher recognition accuracy and faster computational speed of the approach, and illustrated the effectiveness for distinguishing the different loads with promising results. © 2014 IEEE.

In this paper we present our work on human body movement analysis, especially on fall detection. We have developed a reliable dynamic motion pattern analysis algorithm to detect fall situation. The algorithm works on the digital signal output from waist-mounted accelerometry. It first filters noisy components with a Gaussian filter; secondly sets up a 3D body motion model which relates various body postures to the outputs of accelerometry; finally a dynamic detection process is applied to make decision. Experiments were done on 40 cases mimicking various body movements. Our approach gave right judgements in all cases. Our work is an important part of elder care and rehabilitation. ©2004 IEEE.

Knowledge-based image segmentation is defined in terms of the separation of image analysis procedures and representation of knowledge. Such architecture is particularly suitable for medical image segmentation, because of the large amount of structured domain knowledge. A general methodology for the application of knowledge-based methods to medical image segmentation is described. This includes frames for knowledge representation, fuzzy logic for anatomical variations, and a strategy for determining the order of segmentation from the model specification. This method has been applied to three separate problems, 3D thoracic CT, chest X-rays and CT angiography. The application of the same methodology to such a range of applications suggests a major role in medical imaging for segmentation methods incorporating representation of anatomical knowledge.