RSS 1.0Computer Science
Staff within the Department of Computer Science have research interests in Visualization, Interaction & Computer Graphics (with a particular focus on Medical Graphics), Cyber Security and Discrete Optimisation. This collection is licenced under a Creative Commons licence. The collection may be reproduced for non-commerical use and without modification, providing that copyright is acknowledged.
Collections in this community
Recent Submissions
-
Novel ultra-energy-efficient reversible designs of sequential logic quantum-dot cellular automata flip-flop circuitsQuantum-dot cellular automata (QCA) is a technological approach to implement digital circuits with exceptionally high integration density, high switching frequency, and low energy dissipation. QCA circuits are a potential solution to the energy dissipation issues created by shrinking microprocessors with ultra-high integration densities. Current QCA circuit designs are irreversible, yet reversible circuits are known to increase energy efficiency. Thus, the development of reversible QCA circuits will further reduce energy dissipation. This paper presents novel reversible and irreversible sequential QCA set/reset (SR), data (D), Jack Kilby (JK), and toggle (T) flip-flop designs based on the majority gate that utilizes the universal, standard, and efficient (USE) clocking scheme, which allows the implementation of feedback paths and easy routing for sequential QCA-based circuits. The simulation results confirm that the proposed reversible QCA USE sequential flip-flop circuits exhibit energy dissipation less than the Landauer energy limit. Irreversible QCA USE flip-flop designs, although having higher energy dissipation, sometimes have floorplan areas and delay times less than those of reversible designs; therefore, they are also explored. The trade-offs between the energy dissipation versus the area cost and delay time for the reversible and irreversible QCA circuits are examined comprehensively.
-
DNS tunnelling, exfiltration and detection over cloud environmentsThe domain name system (DNS) protocol is fundamental to the operation of the internet, however, in recent years various methodologies have been developed that enable DNS attacks on organisations. In the last few years, the increased use of cloud services by organisations has created further security challenges as cyber criminals use numerous methodologies to exploit cloud services, configurations and the DNS protocol. In this paper, two different DNS tunnelling methods, Iodine and DNScat, have been conducted in the cloud environment (Google and AWS) and positive results of exfiltration have been achieved under different firewall configurations. Detection of malicious use of DNS protocol can be a challenge for organisations with limited cybersecurity support and expertise. In this study, various DNS tunnelling detection techniques were utilised in a cloud environment to create an effective monitoring system with a reliable detection rate, low implementation cost, and ease of use for organisations with limited detection capabilities. The Elastic stack (an open-source framework) was used to configure a DNS monitoring system and to analyse the collected DNS logs. Furthermore, payload and traffic analysis techniques were implemented to identify different tunnelling methods. This cloud-based monitoring system offers various detection techniques that can be used for monitoring DNS activities of any network especially accessible to small organisations. Moreover, the Elastic stack is open-source and it has no limitation with regards to the data that can be uploaded daily.
-
Ret-gadgets in RISC-V-based Binaries Resulting in Traps for HijackersThe presence of instructions within executable programs is what makes the binaries executable. However, attackers leverage on the same to achieve some form of Control Flow Hijacking (CFH). Such code re-use attacks have also been found to lead to Denial of Service (DoS). An example of code re-use attack is Return Oriented Programming (ROP) which is caused by passing input crafted as chained sequences of instructions that are already existing as subroutines in the target program. The instructions are called gadgets and they would normally end with ret. The ret instructions enable the flow of hijacked execution from one set of instruction to another within the attacker’s control. There could however be exceptions depending on the structure of the chained gadgets where the chained gadget fails to run its course due to inability of specific gadgets to replace the value in the return address (ra) register. The dangers of chained gadgets are not a new idea but the possibility for an attacker’s gadget chain to fall into a trap during a ROP attack is not commonly addressed. In addition to this, recent studies have revealed that understanding the behaviours of gadgets would be useful for building information base in training machine learning (ML) models to combat ROP. This study explains the behaviour of certain ROP gadgets showing the possibility of occurrence of a loop in execution during exploitation. A sample program which accesses gadgets from the GNU C library (glibc) is used to demonstrate the findings. Gadgets identified with this possibility are poor for chaining as they do not contain instructions to load or move new values to the ra register and would produce unreliable exploits. This would result in a trap for the chained gadgets instead of arbitrary code execution, and DoS on the path of the user. This implies that the impact that a ROP chain could have on a targeted process does not only rely on the underlying system architecture but also on relies on the structure of the chained gadget. In this paper, the RISC-V architecture is the focus, new gadget finders (scripts are available) are presented, and sample of chained gadgets are analysed on a RISC-V -based binary.
-
Robot companion cats for people at home with dementia: A qualitative case study on companoticsThe use of robot companion pets for people in care homes has been extensively studied. The results are largely positive and suggest that they are valuable in enhancing wellbeing, communication and behavioural aspects. However, there has been little research in people’s own homes, possibly due to the cost and complexity of some of the robot pets currently available. As dementia affects people in different ways, this study explores the effects of a robot cat for people in their own homes, without specifically investigating the effects on a particular symptom. We utilised a case study design to investigate the proposition that various factors influence the impact of a robot cat on the person living with dementia and their carer, including acceptability of the robot pet and acceptance of dementia and its symptoms. The qualitative analysis explores the similarities and differences within the data which were gathered during interviews with people with dementia and their families. This analysis revealed four themes: Distraction, Communication, Acceptance and rejection, and Connecting with the cat and connecting with others. These themes were synthesised into two overarching themes: the effect of the cat on mood and behaviour, and The interaction with the cat. We present the acceptability and impact of the robot cat on symptoms of dementia, with data presented across and within the group of participants. Our analysis suggests that benefits of the robot pet were evident, and although this was a small-scale study, where they were accepted, robot pets provided positive outcomes for the participants and their families.
-
Supervised machine learning for audio emotion recognition: Enhancing film sound design using audio features, regression models and artificial neural networksThe field of Music Emotion Recognition has become and established research sub-domain of Music Information Retrieval. Less attention has been directed towards the counterpart domain of Audio Emotion Recognition, which focuses upon detection of emotional stimuli resulting from non-musical sound. By better understanding how sounds provoke emotional responses in an audience, it may be possible to enhance the work of sound designers. The work in this paper uses the International Affective Digital Sounds set. A total of 76 features are extracted from the sounds, spanning the time and frequency domains. The features are then subjected to an initial analysis to determine what level of similarity exists between pairs of features measured using Pearson’s r correlation coefficient before being used as inputs to a multiple regression model to determine their weighting and relative importance. The features are then used as the input to two machine learning approaches: regression modelling and artificial neural networks in order to determine their ability to predict the emotional dimensions of arousal and valence. It was found that a small number of strong correlations exist between the features and that a greater number of features contribute significantly to the predictive power of emotional valence, rather than arousal. Shallow neural networks perform significantly better than a range of regression models and the best performing networks were able to account for 64.4% of the variance in prediction of arousal and 65.4% in the case of valence. These findings are a major improvement over those encountered in the literature. Several extensions of this research are discussed, including work related to improving data sets as well as the modelling processes.
-
A multi-genre model for music emotion recognition using linear regressorsMaking the link between human emotion and music is challenging. Our aim was to produce an efficient system that emotionally rates songs from multiple genres. To achieve this, we employed a series of online self-report studies, utilising Russell's circumplex model. The first study (n = 44) identified audio features that map to arousal and valence for 20 songs. From this, we constructed a set of linear regressors. The second study (n = 158) measured the efficacy of our system, utilising 40 new songs to create a ground truth. Results show our approach may be effective at emotionally rating music, particularly in the prediction of valence.
-
Evaluating Use of the Doppler Effect to Enhance Auditory AlertsAuditory alerts are an essential part of many multi-modal interaction scenarios, particularly in safety and mission critical settings, such as hospitals and transportation. A variety of strategies can be employed in the design of auditory alerts, often orienting manipulation of volume and pitch parameters. However, manipulations by applying a Doppler effect are under-investigated. A perceptual listening test is conducted (n = 100) using multiple alert sounds that are subjected to a variety of volume, pitch, and Doppler manipulations, with the unaltered sounds serving as a benchmark. Applying a mixed methods approach consisting of inferential statistics and thematic analysis, it is found that decreases in volume and a Doppler simulation of a sound moving away reduce importance and urgency, increase safety, are harder to detect, and are perceived as being more distant in perceptions of auditory alerts. Further, increases in volume and a Doppler simulation of a sound approaching are effective in communicating safety, whilst pitch manipulations were much less effective. Further work is required to provide wider, ecologically valid, verification of these findings, particularly as to how listener detection of Doppler and volume manipulations can be improved.
-
Lossless Compression of Neuromorphic Vision Sensor Data Based on Point Cloud RepresentationVisual information varying over time is typically captured by cameras that acquire data via images (frames) equally spaced in time. Using a different approach, Neuromorphic Vision Sensors (NVSs) are emerging visual capturing devices that only acquire information when changes occur in the scene. This results in major advantages in terms of low power consumption, wide dynamic range, high temporal resolution, and lower data rates than conventional video. Although the acquisition strategy already results in much lower data rates than conventional video, such data can be further compressed. To this end, in this paper we propose a lossless compression strategy based on point cloud compression, inspired by the observation that, by appropriately reporting NVS data in a $(x,y,t)$ tridimensional space, we have a point cloud representation of NVS data. The proposed strategy outperforms the benchmark strategies resulting in a compression ratio up to 30% higher for the considered.
-
Design and Simulation of Reversible Time-Synchronized Quantum-Dot Cellular Automata Combinational Logic Circuits with Ultralow Energy DissipationThe quantum-dot cellular automata (QCA) represent emerging nanotechnology that is poised to supersede the current complementary metal-oxide-semiconductor digital integrated circuit technology. QCA constitutes an extremely promising transistor-less paradigm that can be downscaled to the molecular level, thereby facilitating tera-scale device integration and extremely low energy dissipation. Reversible QCA circuits, which have reversibility sustained down from the logical level to the physical level, can execute computing operations dissipating less energy than the Landauer energy limit (kBTln2). Time synchronization of logic gates is an essential additional requirement, especially in cases involving complex circuits, for ensuring accurate computational results. This paper reports the design and simulation of eight new both logically and physically reversible time-synchronized QCA combinational logic circuits. The new circuit design presented here mitigates the clock delay problems, which are caused by the non-synchronization of logic gate information, via the use of an inherently more symmetric circuit configuration. The simulation results confirm the behaviour of the proposed reversible time-synchronized QCA combinational logic circuits which exhibit ultralow energy dissipation and simultaneously provide accurate computational results.
-
Deep Learning based Human Detection in Privacy-Preserved Surveillance VideosVisual surveillance systems have been improving rapidly over the recent past, becoming more capable and pervasive with incorporation of artificial intelligence. At the same time such surveillance systems are exposing the public to new privacy and security threats. There have been an increasing number of reports of blatant abuse of surveillance technologies. To counteract this, data privacy regulations (e.g. GDPR in Europe) have provided guidelines for data collection and data processing. However, there is still a need for a private and secure method of model training for advanced machine learning and deep learning algorithms. To this end, in this paper we propose a privacy-preserved method for visual surveillance. We first develop a dataset of privacy preserved videos. The data in these videos is masked using Gaussian Mixture Model (GMM) and selective encryption. We then train high-performance object detection models on the generated dataset. The proposed method utilizes state-of-art object detection deep learning models (viz. YOLOv4 and YOLOv5) to perform human/object detection in masked videos. The results are encouraging, and are pointers to the viability of the use of modern day deep learning models for object detection in privacy-preserved videos.
-
Local-Partial Signal Combining Schemes for Cell-Free Large-Scale MU-MIMO Systems with Limited Fronthaul Capacity and Spatial Correlation ChannelsCell-free large-scale multi-user MIMO is a promising technology for the 5G-and-beyond mobile communication networks. Scalable signal processing is the key challenge in achieving the benefits of cell-free systems. This study examines a distributed approach for cell-free deployment with user-centric configuration and finite fronthaul capacity. Moreover, the impact of scaling the pilot length, the number of access points (APs), and the number of antennas per AP on the achievable average spectral efficiency are investigated. Using the dynamic cooperative clustering (DCC) technique and large-scale fading decoding process, we derive an approximation of the signal-tointerference-plus-noise ratio in the criteria of two local combining schemes: Local-Partial Regularized Zero Forcing (RZF) and Local Maximum Ratio (MR). The results indicate that distributed approaches in the cell-free system have the advantage of decreasing the fronthaul signaling and the computing complexity. The results also show that the Local-Partial RZF provides the highest average spectral efficiency among all the distributed combining schemes because the computational complexity of the Local-Partial RZF is independent of the UTs. Therefore, it does not grow as the number of user terminals (UTs) increases.
-
AFOM: Advanced Flow of Motion Detection Algorithm for Dynamic Camera VideosThe surveillance videos taken from dynamic cam-eras are susceptible to multiple security threats like replay attacks, man-in-the-middle attacks, pixel correlation attacks etc. Using unsupervised learning, it is a challenge to detect objects in such surveillance videos, as fixed objects may appear to be in motion alongside the actual moving objects. But despite this challenge, the unsupervised learning techniques are efficient as they save object labelling and model training time, which is usually a case with supervised learning models. This paper proposes an effective computer vision-based object identification algorithm that can detect and separate stationary objects from moving objects in such videos. The proposed Advanced Flow Of Motion (AFOM) algorithm takes advantage of motion estimation between two consecutive frames and induces the estimated motion back to the frame to provide an improved detection on the dynamic camera videos. The comparative analysis demonstrates that the proposed AFOM outperforms a traditional dense optical flow (DOF) algorithm with an average increased difference of 56 % in accuracy, 61 % in precision, and 73 % in pixel space ratio (PSR), and with minimal higher object detection timing.
-
A single-layer asymmetric RNN with low hardware complexity for solving linear equationsA single layer neural network for the solution of linear equations is presented. The proposed circuit is based on the standard Hopfield model albeit with the added flexibility that the interconnection weight matrix need not be symmetric. This results in an asymmetric Hopfield neural network capable of solving linear equations. PSPICE simulation results are given which verify the theoretical predictions. A simple technique to incorporate re-configurability into the circuit for setting the different weights of the interconnection is also included. Experimental results for circuits set up to solve small problems further confirm the operation of the proposed circuit.
-
FireNet-v2: Improved Lightweight Fire Detection Model for Real-Time IoT ApplicationsFire hazards cause huge ecological, social and economical losses in day to day life. Due to the rapid increase in the prevalence of fire accidents, it has become vital to equip the assets with fire prevention systems. There have been numerous researches to build a fire detection model in order to avert such accidents, with recent approaches leveraging the enormous improvements in computer vision deep learning models. However, most deep learning models have to compromise with their performance and accurate detection to maintain a reasonable inference time and parameter count. In this paper, we present a customized lightweight convolution neural network for early detection of fire. By virtue of low parameter count, the proposed model is amenable to embedded applications in real-time fire monitoring equipment, and even upcoming fire monitoring approaches such as unmanned aerial vehicles (drones). The fire detection results show marked improvement over the predecessor low-parameter-count models, while further reducing the number of trainable parameters. The overall accuracy of FireNet-v2, which has only 318,460 parameters, was found to be 98.43% when tested over Foggia's dataset.
-
Deep Learning based Wireless Channel Prediction: 5G ScenarioIn the area of wireless communication, channel estimation is a challenging problem due to the need for real-time implementation as well as system dependence on the estimation accuracy. This work presents a Long-Short Term Memory (LSTM) based deep learning (DL) approach for the prediction of channel response in real-time and real-world non-stationary channels. The model uses the pre-defined history of channel impulse response (CIR) data along with two other features viz. transmitter-receiver update distance and root-mean-square delay spread values which are also changing in time with the channel impulse response. The objective is to obtain an approximate estimate of CIRs using prediction through the DL model as compared to conventional methods. For training the model, a sample dataset is generated through the open-source channel simulation software NYUSIM which realizes samples of CIRs for measurement-based channel models based on various multipath channel parameters. From the model test results, it is found that the proposed DL approach provides a viable lightweight solution for channel prediction in wireless communication.
-
Watchdog Monitoring for Detecting and Handling of Control Flow Hijack on RISC-V-based BinariesAbstract—Control flow hijacking has been a major challenge in software security. Several means of protections have been developed but insecurities persist. This is because existing protections have sometimes been circumvented while some resilient protections do not cover all applications. Studies have revealed that a holistic way of tackling software insecurity could involve watchdog monitoring and detection via Control Flow Integrity (CFI). The CFI concept has shown a good measure of reliability to mitigate control flow hijacking. However, sophisticated attack techniques in the form of Return Oriented Programming (ROP) have persisted. A flexible protection is desirable, which not only covers as many architecture structures as possible but also mitigates known resilient attacks like ROP. The solution proffered here is a hybrid of CFI and watchdog timing via inter-process signaling (IP-CFI). It is a software-based protection that involves recompilation of the target program. The implementation here is on vulnerable RISC-V-based process but is flexible and could be adapted on other architectures. We present a proof of concept in IP-CFI which when applied to a vulnerable program, ROP is mitigated. The target program incurs a run-time overhead of 1.5%. The code is available.
-
An Immersive Haptic-enabled Training Simulation for ParamedicsThis paper describes the integration of haptics support into a virtual reality training simulation aimed at skills retention for paramedics. We focus on a chest decompression, a life-saving invasive procedure used for trauma-associated cardiopulmonary resuscitation (and other causes) that every emergency physician needs to master. It is not regularly performed by a paramedic, however, and therefore skills maintenance is a challenge. In our simulation, a virtual Russell PneumoFix-8 device is used to carry out the procedure and it is controlled with the 3D Systems Touch grounded force feedback device. We describe how this device has been integrated into an immersive virtual environment so that it or any other tool can be used at any location in the scene. Quantitative data has been obtained from an evaluation exercise carried out with 21 paramedics. The majority of these participants reported a good feeling of presence, according to the Spatial Presence Experience Scale. They indicated strongly that the use of haptic-enabled simulators that include the kind of interaction techniques implemented in our simulator would be beneficial for training and skills retention. The realism of using the simulator at a 1 to 1 scale was also highly scored. A System Usability Scale was also calculated and the results show that the simulator is close to an acceptable standard for usability but more work is needed. We will address this in future work.
-
CAP: Patching the Human VulnerabilityCyber threats to organisations across all industries are increasing in both volume and complexity, leading to significant, and sometimes severe, conse-quences. The common weakest link in organisations security is the human vulnerability. The sudden popularity of remote-working due to the Covid-19 pandemic opened organisations and their employees up to more risks, partic-ularly as many workers believe that they are more distracted when at home. Existing cyber training using a ‘one-size-fits-all’ approach has been proven inefficient/ineffective and the need for a more fit-for-purpose training is re-quired. When it comes to cyber training, we know that there is no single-training-fits-all solution – people have different technical skills, different prior knowledge and experience, are in different roles, exposed to different security risks, and require knowledge that is relevant to what they do. This study makes a case for tailored role-based cybersecurity training suitable for awareness within organisations across multiple industries. The study ex-plores the strengths and weaknesses of existing cyber training and literature to make recommendations on efficient awareness and training programme strategies. The study carries out knowledge and task analysis of job roles to create profiles of skills and knowledge they require. These are grouped by topic and level to form scenario-based multiple-choice questions which are mapped to create a Cyber Awareness Platform (CAP). A CAP prototype is in-troduced as a flexible web-based system allowing users to assess their prior knowledge and skills personalised to their role. Knowledge gaps and training needs are identified, and recommendations are tailored to the individual. Ini-tial analysis of CAP shows promising results, indicating that such role-sensitive solution would be highly beneficial to users. This offers further de-velopment opportunities in producing an all-in-one cyber assessment and training platform.
-
Forensic Trails Obfuscation and Preservation via Hard Drive FirmwareThe hard disk drive stores data the user is creating, modifying, and deleting while a firmware facilitates communication between the drive and the operating system. The firmware tells the device and machine how to communicate with each other and will share useful information such as, disk size and information on any bad sectors. Current research shows that exploits exist that can manipulate these outputs. As an attacker, you can change the size of the disk displayed to the operating system to hide data in, likewise by marking an area of the disk as bad. Users may not be aware of these changes as the operating system will accept the readings from the firmware. However, although the data is not reachable via the operating system this paper looks at the traceability of manipulated data using data recovery software FTK Imager, Recuva, EaseUS and FEX Imager. This report examines the use of malicious techniques to thwart digital forensic procedures by manipulating the firmware. It is shown how this is possible and current forensic techniques or software does not easily detect a change within the firmware. However, with the use of various forensic tools, obfuscated trails are detectable. This report follows a black box testing methodology to show the validation of forensic tools or software against anti-forensic techniques. The analysis of the results showed that most tools can find the firmware changes, however, it requires an analyst to spot the subtle differences between standard and manipulated devices. The use of multiple software tools can help an analyst spot the inconsistencies.
-
Visualization for Epidemiological Modelling: Challenges, Solutions, Reflections & RecommendationsWe report on an ongoing collaboration between epidemiological modellers and visualization researchers by documenting and reflecting upon knowledge constructs – a series of ideas, approaches and methods taken from existing visualization research and practice – deployed and developed to support modelling of the COVID-19 pandemic. Structured independent commentary on these efforts is synthesized through iterative reflection to develop: evidence of the effectiveness and value of visualization in this context; open problems upon which the research communities may focus; guidance for future activity of this type; and recommendations to safeguard the achievements and promote, advance, secure and prepare for future collaborations of this kind. In describing and comparing a series of related projects that were undertaken in unprecedented conditions, our hope is that this unique report, and its rich interactive supplementary materials, will guide the scientific community in embracing visualization in its observation, analysis and modelling of data as well as in disseminating findings. Equally we hope to encourage the visualization community to engage with impactful science in addressing its emerging data challenges. If we are successful, this showcase of activity may stimulate mutually beneficial engagement between communities with complementary expertise to address problems of significance in epidemiology and beyond. https://ramp-vis.github.io/RAMPVIS-PhilTransA-Supplement/
