https://www.um.edu.mt/library/oar/handle/123456789/8368 Previously known as Department of Intelligent Computer Systems 2026-09-08T02:16:36Z https://www.um.edu.mt/library/oar/handle/123456789/148963 Title: Dynamic ensembles towards out-of-distribution generalization of affect models Authors: Vella Caruana, Sean; Papathanasiou, Athanasios; Makantasis, Konstantinos Abstract: This paper addresses the challenge of out-of-distributiong eneralization in affect modeling by introducing a novel dynamic ensemble approach. Current affect models face significant limitations in genneralizability due to their reliance on the statistical learning paradigm, which assumes consistency between training and testing data distributions. However, affective data inherently violates this assumption as each user-task-annotator combination generates data from a unique distribution. We reformulate affect modeling as a multi-domain learning task and develop a methodology for constructing dynamic ensemble models that enhance generalization across different domains. Unlike static ensemble approaches, our method combines domain-specific models by dynamically weighing their contributions at the individual data point level based on distributional and prediction properties. Our approach draws inspiration from unsupervised domain adaptation techniques and determines ensemble weights proportional to the likelihood that a given data point originates from a specific domain and the prediction confidence. Experimental validation using the RECOLA dataset demonstrates that our proposed methodology significantly outperforms alternative approaches, including fixed ensemble methods and conventional unsupervised domain adaptation techniques. Notably, our dynamic ensembl emethod is general and can be applied to any multiple domain problem beyond affect modeling tasks, offering a versatile solution for addressing distribution shifts and reducing domain gaps in various machine learning applications. 2025-01-01T00:00:00Z https://www.um.edu.mt/library/oar/handle/123456789/148957 Title: A multimodal causal-inspired approach towards affect modeling Authors: Farrugia, Kevin; Papathanasiou, Athanasios; Makantasis, Konstantinos Abstract: This study introduces a novel causal-inspired approach to affect modeling that leverages distribution shifts and emotional subjectivity as advantages rather than obstacles. Our methodology employs causal graph discovery algorithms to identify relationships between affect measurements and emotional states across different environments, defined by user-task-annotator combinations. We implement constraint-based (PC) and score-based (GES) algorithms to discover environment-specific causal graphs for constructing Bayesian networks for affect prediction. The approach is validated on the RECOLA dataset, demonstrating competitive performance with neural networks and random forests. A key finding reveals a consistent directed edge between arousal and valence nodes, extending Russell’s circumplex theory by identifying conditional dependencies when affect measurements are explicitly observed. This work contributes by integrating causal inference in affect modeling, implementing multiple graph discovery approaches, validating on a benchmark dataset, and establishing groundwork for developing causal models of affect. 2025-01-01T00:00:00Z https://www.um.edu.mt/library/oar/handle/123456789/148950 Title: Learning using privileged information for litter detection Authors: Bartolo, Matthias; Makantasis, Konstantinos; Seychell, Dylan Abstract: As litter pollution continues to rise globally, developing automated tools capable of detecting litter effectively remains a significant challenge. This study presents a novel approach that combines, for the first time, privileged information with deep learning object detection to improve litter detection while maintaining model efficiency. We evaluate our method across five widely used object detection models, addressing challenges such as detecting small litter and objects partially obscured by grass or stones. In addition to this, a key contribution of our work is the introduction of a novel method for generating privileged information for object detection by encoding bounding box annotations as binary masks, which are used to guide and refine the detection process. Through experiments on both withindataset evaluation on the renowned SODA dataset and crossdataset evaluation on the BDW and UAVVaste litter detection datasets, we demonstrate consistent performance improvements across all models. Our approach not only bolsters detection accuracy within the training sets but also generalises well to other litter detection contexts. Crucially, these improvements are achieved without increasing model complexity or adding extra layers, ensuring computational efficiency and scalability. Our results suggest that this methodology offers a practical solution for litter detection, balancing accuracy and efficiency in realworld applications. 2025-01-01T00:00:00Z https://www.um.edu.mt/library/oar/handle/123456789/148548 Title: From measurement to mitigation : exploring the transferability of debiasing approaches to gender bias in Maltese language models Authors: Galea, Melanie; Borg, Claudia Abstract: The advancement of Large Language Models (LLMs) has transformed Natural Language Processing (NLP), enabling performance across diverse tasks with little task-specific training. However, LLMs remain susceptible to social biases, particularly reflecting harmful stereotypes from training data, which can disproportionately affect marginalised communities.We measure gender bias in Maltese LMs, arguing that such bias is harmful as it reinforces societal stereotypes and fails to account for gender diversity, which is especially problematic in gendered, low-resource languages.While bias evaluation and mitigation efforts have progressed for English-centric models, research on low-resourced and morphologically rich languages remains limited. This research investigates the transferability of debiasing methods to Maltese language models, focusing on BERTu and mBERTu, BERT-based monolingual and multilingual models respectively. Bias measurement and mitigation techniques from English are adapted to Maltese, using benchmarks such as CrowS-Pairs and SEAT, alongside debiasing methods Counterfactual Data Augmentation, Dropout Regularization, Auto-Debias, and GuiDebias. We also contribute to future work in the study of gender bias in Maltese by creating evaluation datasets.Our findings highlight the challenges of applying existing bias mitigation methods to linguistically complex languages, underscoring the need for more inclusive approaches in the development of multilingual NLP. 2025-01-01T00:00:00Z