Building safeguards against illegitimate access and authentication is a cornerstone for securing systems. Existing user authentication schemes suffer from challenges in detecting impersonation attacks which leave systems vulnerable and susceptible to misuse. A range of research proposals have suggested continuous multimodal biometric authentication (CMBA) systems as a reliable solution. Though contemporary authentication systems have the potential to change their current authentication scheme, there is a lack of critical analysis of current progress in the field to foster and influence practical solutions. This paper provides a systematic survey of existing literature on CMBA systems, followed by analysis to identify and discuss current research and future trends. The study has found that many diverse biometric characteristics are used for multimodal biometric authentication systems. The majority of the studies in the literature reviewed apply supervised learning approaches as a classification technique, and score level fusion is predominantly used as a fusion model. The review has determined however that there is a lack of comparative analysis on CMBA design in terms of combinations of biometric types (behavioural only, physiological only, or both), machine learning algorithms (unsupervised learning and semi-supervised learning), and fusion models. Most of the studies evaluated a CMBA system’s accuracy functionality, such as False Acceptance Rate (FAR), False Rejection Rate (FRR) and Equal Error Rate (EER). However, security, scalability and usability (user acceptance and satisfaction) are generally not addressed thoroughly even though these are key factors for system success in a real deployment. Furthermore, a CMBA system should be implemented and evaluated extensively on real data without restriction to prove that such systems are feasible.

User authentication is crucial in the digital learning environment to preserve the integrity and reliability of the learning process. Implicit authentication using biometrics has been proposed to improve the user experience while resolving the issues that password dominant authentication faces. Implicit authentication does not require explicit user actions as it is a background process that implicitly acquires a user’s identifying information through sensors embedded within the authenticating devices. To accommodate a variety of user contexts, context-aware implicit authentication has gained attention—especially in the mobile device domain—but it has not been fully explored in digital learning environments. This study is motivated to determine how implicit authentication can observe students’ behaviour without causing disruption to their learning activity. The study provides a structured systematic review of the existing literature to identify and discuss the structure of context-aware continuous implicit authentication systems and future directions. The study found that requirements in the future will be: 1) to consider diverse authenticators to cover all possible user interactions with online learning environments, including coverage of course participants not engaged with online exams; 2) to investigate template adaptation to overcome template ageing issues with biometrics; and 3) to explore evaluation approaches of context-aware implicit authentication systems.

Generative Artificial Intelligence (GenAI) techniques have received massive and widespread industry-disrupting attention because they can create synthetic data, images, text, and other types of content that look very close to real-world examples. Unlike traditional AI that mainly focuses on classification or prediction, GenAI models learn and copy data distributions. This allows new applications such as content creation, data augmentation, and training of machine learning models. In this survey, we organise the discussion around two major themes: general strategies in GenAI and domain-specific adaptation. We develop a mapping approach to help researchers and practitioners find suitable GenAI techniques for different domains. For each domain, we provide detailed discussions about data considerations, real-world case studies, the use of domain-related tools and frameworks, and practical guidance. We also review domain-specific challenges and opportunities to support future research and application work. Our findings give a full view of the changing GenAI field and suggest ways to build more effective, flexible, and strong generative solutions.

Multimodal biometric authentication system has been proposed to mitigate issues inherent in systems that rely on a single biometric trait. Studies on multimodal biometric authentication systems have compared different fusion methods to find an optimal fusion level to combine different biometric traits; however the integration of facial and keystroke biometric samples have not been fully investigated. In addition, some recent studies on biometric authentication systems have shown the importance of adaption where systems can mitigate issues of performance degradation over time in dynamic environments. The application of heterogeneous multimodal contexts remains under explored - this study, therefore first investigates the fusion of facial and keystroke biometric traits by comparing feature and score level fusion methods utilising three virtual datasets:Face with CMU, Face with GREYC and Face with Mobile. The findings demonstrate that score-level fusion performs better in terms of accuracy compared to feature-level fusion. Leveraging this result, the study further applies an adaptation mechanism together with score-level fusion. The study updates thresholds and templates by comparing different strategies with the proposal of a hybrid strategy - Dual Path Adaptation and Positive Gallery Protection Adaptation. Instead of relying only on the threshold or using genuine biometric samples alone to determine adaptation, the hybrid strategy considers both imposter and genuine biometric references. Using information from both sources, the hybrid strategy strengthens adaptation decision making to improve recognition accuracy. The results demonstrate the effectiveness of different adaptation methods across three virtual datasets. Dual Path Adaptation achieved the highest performance with recognition accuracy scores of 0.7136 (Face with CMU), 0.6925 (Face with GREYC), and 0.7862 (Face with Mobile), where scores represent the proportion of correctly identified instances (e.g., 0.7136 = 71.36%). Positive Gallery Protection Adaptation followed with 0.6852, 0.7281, and 0.7422, respectively. Adaptive Dual Thresholds yielded 0.7050, 0.6506, and 0.7289, while the baseline without adaptation scored 0.7123, 0.7157, and 0.7499. These findings highlight the importance of selecting an appropriate adaptation strategy, as an unsuitable approach can degrade performance and compromise security. Therefore, it is crucial to identify a well-balanced adaptation method that enhances robustness without sacrificing security.