John G Breslin

• Professor (Full) at Ollscoil na Gaillimhe – University of Galway

Alzheimer's disease (AD) is a neurodegenerative disorder requiring early diagnosis for effective intervention. This study presents a comprehensive evaluation of classical machine learning models, fine-tuned BERT-based models, and large language models (LLMs) for AD diagnosis using structured data. The comparative analysis demonstrates that fine-tun...

A scene graph is a key image representation in visual reasoning. The generalisability of Scene Graph Generation (SGG) methods is crucial for reliable reasoning and real-world applicability. However, imbalanced training datasets limit this, underrepresenting meaningful visual relationships. Current SGG methods using external knowledge sources face l...

This paper presents a real-time multimodal framework to enhance ICU mortality prediction for heart disease patients by integrating structured data, clinical notes, and big data streaming platforms. The proposed framework comprises two components: an offline multimodal model and a real-time pipeline. In the offline component, we proposed the Transfo...

The application of artificial intelligence (AI) in healthcare has been witnessing an increasing interest. Particularly, federated learning (FL) has become favourable due to its potential for enhancing model quality whilst maintaining data privacy and security. However, the effectiveness of present FL methodologies could underperform under non‐IID c...

In the Aspect-Based Sentiment Analysis (ABSA) domain, the Aspect Sentiment Triplet Extraction (ASTE) task has emerged as a pivotal endeavor, offering insights into nuanced aspects, opinions, and sentiment relationships. This paper introduces “MuSe-CarASTE”, an extensive and meticulously curated dataset purpose-built to propel ASTE advancements with...

The world's food security and biodiversity are seriously threatened by the rising frequency and severity of plant diseases, which are made worse by climate change. Plant diseases must be accurately and promptly detected to reduce these hazards, especially when environmental changes are likely to occur. This paper presents a unique method to enhance...

Retrieval-augmented-generation (RAG) improves content gener- ation via a large-language model (LLM) by attaching contextual data to the task presented to the LLM. We will refer to the LLM agent as an autonomous agent with the capability of executing an RAG-based content generation process with its own LLM model and contextual data. In this paper, w...

Adding color to black-and-white speaker videos automatically is a highly desirable technique. It is an artistic process that requires interactivity with humans for the best results. Many existing automatic video colorization systems provide little opportunity for the user to guide the colorization process. In this work, we introduce a novel automat...

Industry 4.0 (I4.0) knowledge graphs are a common way to represent industrial information models. Conventional SPARQL querying systems require the users to be familiar with the data schema and SPARQL syntax. However, this is often very difficult for many users in industrial production, who have mostly an engineering background, instead of a semanti...

Generative Adversarial Network (GAN) can produce realistic synthetic data which can be used in many applications including training other neural networks. However, an adversarial GAN can intentionally train the generator to not include specific data features. This can be exploited by an attacker as this manipulated GAN generated data will create an...

The integration of heterogeneous and unstructured data in Industry 4.0, poses a significant challenge, particularly with advanced manufacturing techniques. To address this issue, Knowledge Graphs (KGs) have emerged as a pivotal technology, yet their deployment often encounters the problem of incompletion due to data diversity and diverse storage fo...

In MLaaS, DNN models are kept in a server operated by the service provider and inputs to the DNN models are provided by the clients. Such inputs are used to execute the DNN models and classification results are sent back to the client. In MLaaS, the DNN model owner does not reveal the DNN model parameters to the client. MLaaS there are a few trust...

Federated machine learning allows multiple collaborative parties to build a machine learning model in a distributed fashion where the data is distributed among the parties. There can be few restrictions on the participation of such parties in a federate machine learning process. For example, parties may have spatial restrictions such that only a li...

Cryptocurrency, despite the upsurge as a speculative investment, is still a long way from being people’s money. The extreme technicality poses a significant barrier for the general public to adopt it as a medium of exchange. Therefore, simplifying the payment process is a predominant necessity; for example, facilitation in adopting conventional pay...

A large Deep Neural Network (DNN) model usually resides within a server in a machine learning as a service (MLaaS) paradigm. The server providing a classification service may not be secure and provide invalid classifications. Hence verification of DNN models in MLaaS is an important problem. A verification of deep neural networks in an MLaaS paradi...

Combining deep learning and common sense knowledge via neurosymbolic integration is essential for semantically rich scene representation and intuitive visual reasoning. This survey paper delves into data- and knowledge-driven scene repre- sentation and visual reasoning approaches based on deep learning, common sense knowledge and neurosymbolic inte...

Estimating the remaining useful life (RUL) of critical industrial assets is of crucial importance for optimizing maintenance strategies, enabling proactive planning of repair tasks, enhanced reliability, and reduced downtime in prognostic health management (PHM). Deep learning-based data-driven approaches have made RUL prediction a lot better, but...

Cloud computing (CC) offers on‐demand computing and resources to users, and organizations, and is also used in many human‐centric intelligent systems. Attacks in cloud networks cause huge damage to service providers and users. Distributed Denial of Service (DDoS) is one of them that greatly impacts the cloud network. The unavailability of resources...

In the realm of smart manufacturing, predictive maintenance plays a pivotal role in ensuring equipment reliability, minimizing downtime, optimizing costs, and reducing product failure rate by detecting faulty products in the early stage. However, the efficacy of predictive maintenance hinges on the quality of training data employed for predictive m...

While current research predominantly focuses on image-based colorization, the domain of video-based colorization remains relatively unexplored. Most existing video colorization techniques operate on a frame-by-frame basis, often overlooking the critical aspect of temporal coherence between successive frames. This approach can result in inconsistenc...

The financial landscape has undergone a profound shift due to the decentralisation and digitisation brought about by cryptocurrencies. As people begin to reevaluate traditional notions of monetary ownership, cryptocurrencies will be poised to play an even more critical role in the future. However, the widespread adoption of cryptocurrencies has bee...

Exploring the potential of neuro-symbolic hybrid approaches offers promising avenues for seamless high-level understanding and reasoning about visual scenes. Scene Graph Generation (SGG) is a symbolic image representation approach based on deep neural networks (DNN) that involves predicting objects, their attributes, and pairwise visual relationshi...

The emergence of Industry 4.0 has transformed modern-day factories into high-tech industrial sites through rapid automation and increased access to real-time data. Deep learning approaches possessing superior capabilities for intelligent, data-driven fault diagnosis have become critical in ensuring process safety and reliability in these industrial...

Peer-to-peer (P2P) energy trading is one of the most effective methods to increase the usage of Renewable Energy (RE) resources in the distribution network and reduce losses by eliminating long transmission and distribution lines. This research aims to enhance the efficiency of P2P energy trading by examining the suitability of four distinct double...

Systems for monitoring air quality are essential for reducing the negative consequences of air pollution, but creating real-time systems encounters several challenges. The accuracy and effectiveness of these systems can be greatly improved by integrating federated learning and multi-access edge computing (MEC) technology. This paper critically revi...

The Internet of Things (IoT) based smart city applications are the latest technology-driven solutions designed to collect and analyze data to enhance the quality of life for urban residents by creating more sustainable, efficient, and connected communities. Communication nodes are networked independently to monitor the circumstances, where they req...

Industry 4.0 (I4.0) is a new era in the industrial revolution that emphasizes machine connectivity, automation, and data analytics. The I4.0 pillars such as autonomous robots, cloud computing, horizontal and vertical system integration, and the industrial internet of things have increased the performance and efficiency of production lines in the ma...

The electric vehicles (EVs) charging stations (CSs) at public premises have higher installation and power consumption costs. The potential benefits of public CSs rely on their efficient utilization. However, the conventional charging methods obligate a long waiting time and thereby deteriorate their efficiency with low utilization. This paper sugge...

Meaningful feature extraction from multivariate time-series data is still challenging since it takes into account the correlation between pairs of sensors as well as the temporal information of each time-series. Meanwhile, the huge industrial system has evolved into a data-rich environment, resulting in the rapid development and deployment of deep...

The current challenging issue in the Advanced Metering Infrastructure (AMI) of the Smart Grid (SG) network is how to classify and identify smart meters under the effect of falsification attacks related to stealthy data, which are injected in such small numbers or percentages that it becomes hard to identify. The problem becomes challenging due to s...

At least one-third of the food produced for human consumption is lost and wasted due to inefficient perishable product distribution plans. Currently used distribution plans are static and centralised as a single entity decides on a distribution plan where products may move along a predefined path. In this paper, we developed a dynamic distribution...

Payment channel networks (PCN) are one of the promising second layer solutions for the scalability issue of cryptocurrencies. They offer high throughput and low-cost transactions for the users. However, practical and design issues of PCN hinder users from embracing the technology. This paper focuses on how to increase user adoption by boosting reve...

Although it is difficult to overwrite the data kept in blockchains, there are numerous incidents of false data insertion in blockchains. Current blockchain technologies can not prevent such false insertion into blockchains. In this paper, we present a blockchain model that can prevent such immutable lies. We have several contributions in this paper...

Blockchain is a promising tool to implement peer-to-peer energy trade algorithms because it lowers the cost of electricity by eliminating 3rd parties such as the utility companies from energy trade and by creating a secure trade platform. However, the state of the art blockchain-based peer to peer energy trade solutions have privacy and scalability...

The ongoing industrial revolution termed Industry 4.0 (I4.0) has borne witness to a series of profound changes towards increasing smart automation, particularly in the industrial sectors of automotive, aerospace, manufacturing, etc. Automatic welding, a widely applied manufacturing process in these domains, is not an exception of these changes. One...

With the advent of Industry 4.0 (I4.0) leading to the proliferation of industrial process data, deep learning (DL) techniques have become instrumental in developing intelligent fault diagnosis (FD) applications. However, de- spite their potentially superior process monitoring capa- bilities, DL-based FD models are poorly calibrated and generate poi...

In recent years Cyber-Physical Systems (CPS) and Industrial Internet of Things (IIoT) have gained significant attraction; however, it remains a vulnerable target for cyberattacks. Machine learning techniques have garnered interest in security applications due to their rapid processing capabilities and real-time predictions. However, imbalanced data...

5G emerges as the bedrock for the Industrial Internet of Things (IIoT), it facilitates the seamless, low-latency fusion of artificial intelligence and cloud computing, thereby fortifying the entire industrial procedure within a framework of smart and intelligent IIoT ecosystems. Concurrently, the continuously changing landscape of cybersecurity thr...

Smart Factories characterize as context-rich, fast-changing environments where heterogeneous hardware appliances are found beside of also heterogeneous software components deployed in (or directly interfacing with) IoT devices, as well as in on-premise mainframes, and on the Cloud. This inherent heterogeneity poses major challenges particularly whe...

TinyML: Tiny in size, big in impact! In this paper, we present a Tensor Memory Mapping (TMM) method, which can accurately calculate the on-device execution memory consumed by a range of ML and TinyML models during execution on small central processing units (CPUs), microcontroller units (MCUs), and single board computers (SBCs).

Motivation. Industry 4.0 [1, 2] comes with unprecedented amounts of heterogeneous industrial data [3,4,5].

Traditionally, original equipment manufacturers (OEMs) send device-specific over-the-air (OTA) packages to ensure the latest firmware, security patches, etc. With millions of IoT devices, even a tiny percentage of OTA failures will result in tens of thousands of globally affected consumers. The state-of-the-art OTA methods are suited for high-end A...

With the introduction of ultra-low-power machine learning (TinyML), IoT devices are becoming smarter as they are driven by ML models. However, any loss of communication at the device level can lead to a failure of the entire IoT system or misleading information transmission. Since there exist numerous heterogeneous devices within an IoT system, it...

Invariable of the agriculture type (precision, smart, or digital), the monitoring process of factors that increase the crop yield and growth is mostly non-ML, manually structured approaches with practical pain points. In this scenario, to reduce monitoring costs and maintenance efforts, there is a requirement for low-cost semi-autonomous distribute...

A correct reputation management system can differentiate between low-quality and high-quality data providers in an IoT data marketplace. There are challenges in designing an unbiased and secure reputation management system that can not be manipulated by wrong feedbacks or wrong aggregation of feedbacks. In this paper, we develop a decentralised rep...

Payment channel networks (PCN) are one of the promising second layer solutions for the scalability issue of cryptocurrencies. They offer high throughput and low-cost transactions for the users. However, practical and design issues of PCN hinder users from embracing the technology. This paper focuses on how to increase user adoption by boosting reve...

Due to increasing concern about climate change, the local energy market has been revolutionized with the increase in solar photovoltaic, electric vehicles, smart home appliances, and demand response. These technologies used in the residential sector provides new opportunities for Home Energy Management System (HEMS) to manage peak hours and gain in...

Scene graph generation aims to capture the semantic elements in images by modelling objects and their relationships in a structured manner, which are essential for visual understanding and reasoning tasks including image captioning, visual question answering, multimedia event processing, visual storytelling and image retrieval. The existing scene g...

Although it is difficult to overwrite the data kept in blockchains, there are numerous incidents of false data insertion in blockchains. Current blockchain technologies can not prevent such false insertion into blockchains. In this paper, we present a blockchain model that can prevent such immutable lies. We have several contributions in this paper...

At least one-third of the food produced for human consumption is lost and wasted due to inefficient perishable product distribution plans. Currently used distribution plans are static and centralised as a single entity decides on a distribution plan where products may move along a predefined path. In this paper, we developed a dynamic distribution...

Blockchain is a promising tool to implement peer-to-peer energy trade algorithms because it lowers the cost of electricity by eliminating 3rd parties such as the utility companies from energy trade and by creating a secure trade platform. However, the state of the art blockchain-based peer to peer energy trade solutions have privacy and scalability...

Peer-to-Peer (P2P) energy trading, one of the new paradigms driven by decentralization, decarbonization, and digitalization of the smart grid, has become a widespread technique in recent years. Additionally, the rise of the double auction for P2P energy trading suggests better trading algorithms with unprecedented economic and technical benefits. T...

Huge advances in peer-to-peer systems and attempts to develop the semantic web have revealed a critical issue in information systems across multiple domains: the absence of semantic interoperability. Today, businesses operating in a digital environment require increased supply-chain automation, interoperability, and data governance. While research...

Visual understanding involves detecting objects in a scene and investigating rich semantic relationships between the objects, which is required for downstream visual reasoning tasks. Scene graph is widely used for structured scene representation, however, the performance of scene graph generation for visual reasoning is limited due to challenges po...

This work presents a two-layer decentralized charging approach (TLDCA) based on fuzzy data fusion concerning the economic and power layers for optimizing the charging cost of residential electric vehicles (EVs). We defined the problem with the fuzzy objective function of minimizing the charging costs and presented a detailed fuzzy integer linear pr...

Scene graph generation aims to capture the semantic elements in images by modelling objects and their relationships in a structured manner, which are essential for visual understanding and reasoning tasks including image captioning, visual question answering, multimedia event processing, visual storytelling and image retrieval. The existing scene g...

This article presents a novel over-the-air (OTA) technique to remotely deploy tiny ML models over Internet of Things (IoT) devices and perform tasks, such as machine learning (ML) model updates, firmware reflashing, reconfiguration, or repurposing. We discuss relevant challenges for OTA ML deployment over IoT both at the scientific and engineering...

In training distributed machine learning, communicating model updates among workers has always been a bottleneck. The magnitude of impact on the quality of resultant models is higher when distributed training on low hardware specification devices and in uncertain real-world IoT networks where congestion, latency, band-width issues are common. In th...

The proliferation of sensing technologies such as sensors has resulted in vast amounts of time-series data being produced by machines in industrial plants and factories. There is much information available that can be used to predict machine breakdown and degradation in a given factory. The downtime of industrial equipment accounts for heavy losses...

The majority of IoT devices like smartwatches, smart plugs, HVAC controllers, etc., are powered by hardware with a constrained specification (low memory, clock speed and processor) which is insufficient to accommodate and execute large, high-quality models. On such resource-constrained devices, manufacturers still manage to provide attractive funct...

TinyML refers to the intersection of machine learning (ML), mathematical optimization, and tiny IoT embedded systems. In the current era of ubiquitous connectivity and pervasive data, TinyML has emerged as an effective method to continuously analyze real-world data without the resource overhead of traditional ML hardware. However, as valuable data...

The charging loads of electric vehicles (EVs) at residential premises are controlled through a tariff system based on fixed timing. The conventional tariff system presents the herding issue, such as with many connected EVs, all of them are directed to charge during the same off-peak period, which results in overloading the power grid and high charg...

Electric vehicles (EVs) need to be recharged at intermediate locations, such as shopping malls, restaurants, and parking lots, to meet the daily commute requirements of their users. Currently, public electric vehicle supply equipment (EVSE) serve EVs by conventional methods, which can result in long waiting time for users. This issue reduces the tr...

Involving an intermediary between producer-consumer environments is a common practice that reduces the managerial overhead on both parties. However, this mediation has both pros and cons. For example, it can overtake the power of producers in pricing, which could cause unfair revenue distribution. We present a novel micropayment protocol called MAR...

Businesses operating in a digital world require increased automation, interoperability, and data governance for supply chain activities. A substantial amount of data generated by these operations remains unused, disconnected, or latent. The purpose of this research is to demonstrate how semantic web development can leverage ontologies to optimise d...

The energy sector is undergoing a paradigm shift to integrate the increasing volume of embedded renewable energy generation and create Local Energy Communities (LEC). Peer-to-Peer (P2P) energy trading is an encouraging paradigm used to increase usage of renewable energy, decrease consumers’ electricity bills, and provide revenue to prosumers. It al...

The concept of ML model aggregation rather than data aggregation has gained much attention as it boosts prediction performance while maintaining stability and preserving privacy. In a non-ideal scenario, there are chances for a base model trained on a single device to make independent but complementary errors. To handle such cases, in this paper, w...

Typically a Neural Networks (NN) is trained on data centers using historic datasets, then a C source file (model as a char array) of the trained model is generated and flashed on IoT devices. This standard process impedes the flexibility of billions of deployed ML-powered devices as they cannot learn unseen/fresh data patterns (static intelligence)...

Social bots can cause social, political, and economical disruptions by spreading rumours. The state of the art methods to prevent social bots from spreading rumours are centralised and such solutions may not be ac- cepted by users who may not trust a centralised solution being biased. In this paper, we developed a decentralised method to prevent so...

Transmitting updates of high-dimensional neural network (NN) models between client IoT devices and the central aggregating server has always been a bottleneck in collaborative learning - especially in uncertain real-world IoT networks where congestion, latency, bandwidth issues are common. In this scenario, gradient quantization is an effective way...

Over the recent years, Industry 4.0 (I4.0) technologies such as the Industrial Internet of Things (IIoT), Artificial Intelligence (AI), and the presence of Industrial Big Data (IBD) have helped achieve intelligent Fault Detection (FD) in manufacturing. Notably, data-driven approaches in FD apply Deep Learning (DL) techniques to help generate insigh...

Online social network users suffer from rumors and misinformation. Most of these networks are centralized and controlled by a single entity. Such centralized entities may intervene to prevent the propagation of social media content that it considers as misinformation. However, such vetting entities may be biased and may selectively prevent misinfor...

Edge analytics refers to the application of data analytics and Machine Learning (ML) algorithms on IoT devices. The concept of edge analytics is gaining popularity due to its ability to perform AI-based analytics at the device level, enabling autonomous decision-making, without depending on the cloud. However, the majority of Internet of Things (Io...

Edge analytics refers to the application of data analytics and Machine Learning (ML) algorithms on IoT devices. The concept of edge analytics is gaining popularity due to its ability to perform AI-based ana-lytics at the device level, enabling autonomous decision-making, without depending on the cloud. However, the majority of Internet of Things (I...

With the introduction of ultra-low-power machine learning (TinyML), IoT devices are becoming smarter as they are driven by Machine Learning (ML) models. However, any increase in the training data results in a linear increase in the space complexity of the ML models. It is highly challenging to deploy such ML models on IoT devices with limited memor...

Monitoring and analyzing air quality is of primary importance to encourage more sustainable lifestyles and plan corrective actions. This paper presents the design and end-to-end implementation of a real-world urban air quality data collection and analytics use case which is a part of the TRAFAIR (Understanding Traffic Flows to Improve Air Quality)...

The development of the Internet of Things (IoT) technology and their integration in smart cities have changed the way we work and live, and enriched our society. However, IoT technologies present several challenges such as increases in energy consumption, and produces toxic pollution as well as E-waste in smart cities. Smart city applications must...

The majority of IoT edge devices are embedded systems with a tiny microcontroller unit (MCU), which acts as its brain. When users want their edge devices to continuously improve for better edge-analytics results, there is a need to equip their devices with algorithms that can learn/train from the continuously evolving real-world data. Currently, su...

Traditional grid network has played a major role in society by distributing and transmitting electric supply to consumers. However, with the advancement in technology in Industry 4.0 has evolved the role of the Smart Grid (SG) network. SG network is a two-way bi-directional communication Cyber-Physical System (CPS). Whereas traditional grid network...

Involving an intermediary between producer-consumer environments is a common practice that reduces the managerial overhead on both parties. However, this mediation has both pros and cons. For example, it can overtake the power of producers in pricing, which could cause unfair revenue distribution. We present a novel micropayment protocol called MAR...

The healthcare Internet-of-Things (IoT) offers many benefits including data transmission in real-time mode, the ability to monitor the physiological state of the patient in a different interval of time. Devices such as blood-pressure monitors, glucose meters, heart monitoring implants, Electroencephalography (EEG), Electrocardiogram (ECG), and Elec...