arXiv subjects

Gelareh Mohammadi

Publications and source records attributed to Gelareh Mohammadi.

At least 19 recordsLinked to original sources

arXiv2026

BARRAC: Adaptation of an English Aspect-based Sentiment Analysis Approach for Classification Tasks in Arabic Dialects

With the rapid growth of Arabic NLP, several models, datasets and benchmarks have been reported. This paper asks whether approaches developed for majority languages like English can be adapted to Arabic tasks. We adapt an English aspect-based sentiment analysis framework to Arabic classification tasks and present the adaptation as BARRAC: Brainstorming Alignment and Replaced Representation learning for ArabiC tasks. BARRAC replaces consumer-review attribute pools with Arabic linguistic devices and markers for dialectal sentiment, sarcasm, and dialect identification, and replaces noisy self-training with two-stage training. Evaluated on five Arabic dialect datasets, BARRAC achieves a mean macro-F1 of 63.93\%, outperforming the best few-label SOTA by 3\%, and outperforming GPT-4o on four out of five tasks. Error analysis provides insights into remaining challenges. These results demonstrate that adapting task-specific approaches is a promising direction for Arabic NLP alongside adapting models, datasets and benchmarks.

cs.CL↗

arXiv2026

LEGOUI: Designing with UI-DSL Bricks to Balance Transparency and Controllability

Generative user interface design tools enable rapid prototyping but often operate as black boxes with limited transparency and controllability. When outputs diverge from the designer's intent, users are left tweaking prompts via trial-and-error with little insight into the model's reasoning. We present LegoUI, a staged generative framework that structures the interface design process into sequential, interpretable steps along key design dimensions, capturing each step's result in a UI domain-specific language (UI-DSL) enriched with provenance. This approach exposes the model's intermediate reasoning and enables user intervention and iterative refinement. In a technical evaluation on 40 real-world design prompts, LegoUI's requirement analysis stage captured explicit requirements with over 95% accuracy, near-complete coverage, and zero redundancy. In user studies, participants using LegoUI reported significantly greater transparency, controllability, and alignment with their intent compared to existing one-shot generative UI tools.

cs.HC↗

arXiv2026

Flick: Few Labels Text Classification using K-Aware Intermediate Learning in Multi-Task Low-Resource Languages

Training deep learning networks with minimal supervision has gained significant research attention due to its potential to reduce reliance on extensive labelled data. While self-training methods have proven effective in semi-supervised learning, they remain vulnerable to errors from noisy pseudo labels. Moreover, most recent approaches to the few-label classification problem are either designed for resource-rich languages such as English or involve complex cascading models that are prone to overfitting. To address the persistent challenge of few-label text classification in truly low-resource linguistic contexts, where existing methods often struggle with noisy pseudo-labels and domain adaptation, we propose Flick. Unlike prior methods that rely on generic multi-cluster pseudo-labelling or complex cascading architectures, Flick leverages the fundamental insight that distilling high-confidence pseudo-labels from a broader set of initial clusters can dramatically improve pseudo-label quality, particularly for linguistically diverse, low-resource settings. Flick introduces a novel pseudo-label refinement component, a departure from traditional pseudo-labelling strategies by identifying and leveraging top-performing pseudo-label clusters. This component specifically learns to distil highly reliable pseudo-labels from an initial broad set by focusing on single-cluster cohesion and leveraging an adaptive top-k selection mechanism. This targeted refinement process is crucial for mitigating the propagation of errors inherent in low-resource data, allowing for robust fine-tuning of pre-trained language models with only a handful of true labels. We demonstrate Flick's efficacy across 14 diverse datasets, encompassing challenging low-resource languages such as Arabic, Urdu, and Setswana, alongside English, showcasing its superior performance and adaptability.

cs.CL↗

arXiv2025

Towards Human-AI Synergy in UI Design: Supporting Iterative Generation with LLMs

In automated UI design generation, a key challenge is the lack of support for iterative processes, as most systems focus solely on end-to-end output. This stems from limited capabilities in interpreting design intent and a lack of transparency for refining intermediate results. To better understand these challenges, we conducted a formative study that identified concrete and actionable requirements for supporting iterative design with Generative Tools. Guided by these findings, we propose PrototypeFlow, a human-centered system for automated UI generation that leverages multi-modal inputs and models. PrototypeFlow takes natural language descriptions and layout preferences as input to generate the high-fidelity UI design. At its core is a theme design module that clarifies implicit design intent through prompt enhancement and orchestrates sub-modules for component-level generation. Designers retain full control over inputs, intermediate results, and final prototypes, enabling flexible and targeted refinement by steering generation and directly editing outputs. Our experiments and user studies confirmed the effectiveness and usefulness of our proposed PrototypeFlow.

cs.HC↗

arXiv2025

GazeDETR: Gaze Detection using Disentangled Head and Gaze Representations

Gaze communication plays a crucial role in daily social interactions. Quantifying this behavior can help in human-computer interaction and digital phenotyping. While end-to-end models exist for gaze target detection, they only utilize a single decoder to simultaneously localize human heads and predict their corresponding gaze (e.g., 2D points or heatmap) in a scene. This multitask learning approach generates a unified and entangled representation for human head localization and gaze location prediction. Herein, we propose GazeDETR, a novel end-to-end architecture with two disentangled decoders that individually learn unique representations and effectively utilize coherent attentive fields for each subtask. More specifically, we demonstrate that its human head predictor utilizes local information, while its gaze decoder incorporates both local and global information. Our proposed architecture achieves state-of-the-art results on the GazeFollow, VideoAttentionTarget and ChildPlay datasets. It outperforms existing end-to-end models with a notable margin.

cs.CV↗

arXiv2025

A Case Study of Scalable Content Annotation Using Multi-LLM Consensus and Human Review

Content annotation at scale remains challenging, requiring substantial human expertise and effort. This paper presents a case study in code documentation analysis, where we explore the balance between automation efficiency and annotation accuracy. We present MCHR (Multi-LLM Consensus with Human Review), a novel semi-automated framework that enhances annotation scalability through the systematic integration of multiple LLMs and targeted human review. Our framework introduces a structured consensus-building mechanism among LLMs and an adaptive review protocol that strategically engages human expertise. Through our case study, we demonstrate that MCHR reduces annotation time by 32% to 100% compared to manual annotation while maintaining high accuracy (85.5% to 98%) across different difficulty levels, from basic binary classification to challenging open-set scenarios.

cs.HC↗

arXiv2024

DesignRepair: Dual-Stream Design Guideline-Aware Frontend Repair with Large Language Models

The rise of Large Language Models (LLMs) has streamlined frontend interface creation through tools like Vercel's V0, yet surfaced challenges in design quality (e.g., accessibility, and usability). Current solutions, often limited by their focus, generalisability, or data dependency, fall short in addressing these complexities. Moreover, none of them examine the quality of LLM-generated UI design. In this work, we introduce DesignRepair, a novel dual-stream design guideline-aware system to examine and repair the UI design quality issues from both code aspect and rendered page aspect. We utilised the mature and popular Material Design as our knowledge base to guide this process. Specifically, we first constructed a comprehensive knowledge base encoding Google's Material Design principles into low-level component knowledge base and high-level system design knowledge base. After that, DesignRepair employs a LLM for the extraction of key components and utilizes the Playwright tool for precise page analysis, aligning these with the established knowledge bases. Finally, we integrate Retrieval-Augmented Generation with state-of-the-art LLMs like GPT-4 to holistically refine and repair frontend code through a strategic divide and conquer approach. Our extensive evaluations validated the efficacy and utility of our approach, demonstrating significant enhancements in adherence to design guidelines, accessibility, and user experience metrics.

cs.SE↗

arXiv2024

Investigating the Design Considerations for Integrating Text-to-Image Generative AI within Augmented Reality Environments

Generative Artificial Intelligence (GenAI) has emerged as a fundamental component of intelligent interactive systems, enabling the automatic generation of multimodal media content. The continuous enhancement in the quality of Artificial Intelligence-Generated Content (AIGC), including but not limited to images and text, is forging new paradigms for its application, particularly within the domain of Augmented Reality (AR). Nevertheless, the application of GenAI within the AR design process remains opaque. This paper aims to articulate a design space encapsulating a series of criteria and a prototypical process to aid practitioners in assessing the aptness of adopting pertinent technologies. The proposed model has been formulated based on a synthesis of design insights garnered from ten experts, obtained through focus group interviews. Leveraging these initial insights, we delineate potential applications of GenAI in AR.

cs.HC↗

arXiv2024

Exploring Emotions in Multi-componential Space using Interactive VR Games

Emotion understanding is a complex process that involves multiple components. The ability to recognise emotions not only leads to new context awareness methods but also enhances system interaction's effectiveness by perceiving and expressing emotions. Despite the attention to discrete and dimensional models, neuroscientific evidence supports those emotions as being complex and multi-faceted. One framework that resonated well with such findings is the Component Process Model (CPM), a theory that considers the complexity of emotions with five interconnected components: appraisal, expression, motivation, physiology and feeling. However, the relationship between CPM and discrete emotions has not yet been fully explored. Therefore, to better understand emotions underlying processes, we operationalised a data-driven approach using interactive Virtual Reality (VR) games and collected multimodal measures (self-reports, physiological and facial signals) from 39 participants. We used Machine Learning (ML) methods to identify the unique contributions of each component to emotion differentiation. Our results showed the role of different components in emotion differentiation, with the model including all components demonstrating the most significant contribution. Moreover, we found that at least five dimensions are needed to represent the variation of emotions in our dataset. These findings also have implications for using VR environments in emotion research and highlight the role of physiological signals in emotion recognition within such environments.

cs.HC↗

arXiv2024

A critical analysis of cognitive load measurement methods for evaluating the usability of different types of interfaces: guidelines and framework for Human-Computer Interaction

Usability testing is an essential part of product design, particularly for user interfaces. To enhance the reliability of usability evaluations, employing cognitive load measurement methods can be highly effective in assessing the mental effort required to complete tasks during user testing. This review aims to provide an overview of the most suitable cognitive load measurement methods for evaluating various types of user interfaces, serving as a valuable resource for guiding usability assessments. To bridge the existing gap in the literature, a systematic review was conducted, analyzing 76 articles with experimental study designs that met the eligibility criteria. The review encompasses different methods of measuring cognitive load applicable to assessing the usability of diverse user interfaces, including computer software, information systems, video games, web and mobile applications, robotics, and virtual reality applications. The results highlight the most widely utilized cognitive load measurement methods in software usability, their respective usage percentages, and their application in evaluating the usability of each user interface type. Additionally, the advantages and disadvantages of each method are discussed. Furthermore, the review proposes a framework to assist usability testers in selecting an appropriate cognitive load measurement method for conducting accurate usability evaluations.

cs.HC↗

arXiv2023

Speech-Gesture GAN: Gesture Generation for Robots and Embodied Agents

Embodied agents, in the form of virtual agents or social robots, are rapidly becoming more widespread. In human-human interactions, humans use nonverbal behaviours to convey their attitudes, feelings, and intentions. Therefore, this capability is also required for embodied agents in order to enhance the quality and effectiveness of their interactions with humans. In this paper, we propose a novel framework that can generate sequences of joint angles from the speech text and speech audio utterances. Based on a conditional Generative Adversarial Network (GAN), our proposed neural network model learns the relationships between the co-speech gestures and both semantic and acoustic features from the speech input. In order to train our neural network model, we employ a public dataset containing co-speech gestures with corresponding speech audio utterances, which were captured from a single male native English speaker. The results from both objective and subjective evaluations demonstrate the efficacy of our gesture-generation framework for Robots and Embodied Agents.

cs.AI↗

arXiv2023

Effective permeability of an immiscible fluid in porous media determined from its geometric state

Based on the phenomenological extension of Darcy's law, two-fluid flow is dependent on a relative permeability function of saturation only that is process/path dependent with an underlying dependency on pore structure. For applications, fuel cells to underground $CO_2$ storage, it is imperative to determine the effective phase permeability relationships where the traditional approach is based on the inverse modelling of time-consuming experiments. The underlying reason is that the fundamental upscaling step from pore to Darcy scale, which links the pore structure of the porous medium to the continuum hydraulic conductivities, is not solved. Herein, we develop an Artificial Neural Network (ANN) that relies on fundamental geometrical relationships to determine the mechanical energy dissipation during creeping immiscible two-fluid flow. The developed ANN is based on a prescribed set of state variables based on physical insights that predicts the effective permeability of 4,500 unseen pore-scale geometrical states with $R^2 = 0.98$.

physics.flu-dyn↗

arXiv2023

EmoStim: A Database of Emotional Film Clips with Discrete and Componential Assessment

Emotion elicitation using emotional film clips is one of the most common and ecologically valid methods in Affective Computing. However, selecting and validating appropriate materials that evoke a range of emotions is challenging. Here we present EmoStim: A Database of Emotional Film Clips as a film library with a rich and varied content. EmoStim is designed for researchers interested in studying emotions in relation to either discrete or componential models of emotion. To create the database, 139 film clips were selected from literature and then annotated by 638 participants through the CrowdFlower platform. We selected 99 film clips based on the distribution of subjective ratings that effectively distinguished between emotions defined by the discrete model. We show that the selected film clips reliably induce a range of specific emotions according to the discrete model. Further, we describe relationships between emotions, emotion organization in the componential space, and underlying dimensions representing emotional experience. The EmoStim database and participant annotations are freely available for research purposes. The database can be used to enrich our understanding of emotions further and serve as a guide to select or create additional materials.

cs.HC↗

arXiv2022

2CET-GAN: Pixel-Level GAN Model for Human Facial Expression Transfer

Recent studies have used GAN to transfer expressions between human faces. However, existing models have many flaws: relying on emotion labels, lacking continuous expressions, and failing to capture the expression details. To address these limitations, we propose a novel CycleGAN- and InfoGAN-based network called 2 Cycles Expression Transfer GAN (2CET-GAN), which can learn continuous expression transfer without using emotion labels. The experiment shows our network can generate diverse and high-quality expressions and can generalize to unknown identities. To the best of our knowledge, we are among the first to successfully use an unsupervised approach to disentangle expression representation from identities at the pixel level.

cs.CV↗

arXiv2022

Temporal Pattern Mining for Analysis of Longitudinal Clinical Data: Identifying Risk Factors for Alzheimer's Disease

A novel framework is proposed for handling the complex task of modelling and analysis of longitudinal, multivariate, heterogeneous clinical data. This method uses temporal abstraction to convert the data into a more appropriate form for modelling, temporal pattern mining, to discover patterns in the complex, longitudinal data and machine learning models of survival analysis to select the discovered patterns. The method is applied to a real-world study of Alzheimer's disease (AD), a progressive neurodegenerative disease that has no cure. The patterns discovered were predictive of AD in survival analysis models with a Concordance index of up to 0.8. This is the first work that performs survival analysis of AD data using temporal data collections for AD. A visualisation module also provides a clear picture of the discovered patterns for ease of interpretability.

cs.LG↗

arXiv2022

Ensemble feature selection with clustering for analysis of high-dimensional, correlated clinical data in the search for Alzheimer's disease biomarkers

Healthcare datasets often contain groups of highly correlated features, such as features from the same biological system. When feature selection is applied to these datasets to identify the most important features, the biases inherent in some multivariate feature selectors due to correlated features make it difficult for these methods to distinguish between the important and irrelevant features and the results of the feature selection process can be unstable. Feature selection ensembles, which aggregate the results of multiple individual base feature selectors, have been investigated as a means of stabilising feature selection results, but do not address the problem of correlated features. We present a novel framework to create feature selection ensembles from multivariate feature selectors while taking into account the biases produced by groups of correlated features, using agglomerative hierarchical clustering in a pre-processing step. These methods were applied to two real-world datasets from studies of Alzheimer's disease (AD), a progressive neurodegenerative disease that has no cure and is not yet fully understood. Our results show a marked improvement in the stability of features selected over the models without clustering, and the features selected by these models are in keeping with the findings in the AD literature.

cs.LG↗

arXiv2022

Ensemble feature selection with data-driven thresholding for Alzheimer's disease biomarker discovery

Healthcare datasets present many challenges to both machine learning and statistics as their data are typically heterogeneous, censored, high-dimensional and have missing information. Feature selection is often used to identify the important features but can produce unstable results when applied to high-dimensional data, selecting a different set of features on each iteration. The stability of feature selection can be improved with the use of feature selection ensembles, which aggregate the results of multiple base feature selectors. A threshold must be applied to the final aggregated feature set to separate the relevant features from the redundant ones. A fixed threshold, which is typically applied, offers no guarantee that the final set of selected features contains only relevant features. This work develops several data-driven thresholds to automatically identify the relevant features in an ensemble feature selector and evaluates their predictive accuracy and stability. To demonstrate the applicability of these methods to clinical data, they are applied to data from two real-world Alzheimer's disease (AD) studies. AD is a progressive neurodegenerative disease with no known cure, that begins at least 2-3 decades before overt symptoms appear, presenting an opportunity for researchers to identify early biomarkers that might identify patients at risk of developing AD. Features identified by applying these methods to both datasets reflect current findings in the AD literature.

cs.LG↗

arXiv2021

Virtual Reality for Emotion Elicitation -- A Review

Emotions are multifaceted phenomena that affect our behaviour, perception, and cognition. Increasing evidence indicates that induction mechanisms play a crucial role in triggering emotions by simulating the sensations required for an experimental design. Over the years, many reviews have evaluated a passive elicitation mechanism where the user is an observer, ignoring the importance of self-relevance in emotional experience. So, in response to the gap in the literature, this study intends to explore the possibility of using Virtual Reality (VR) as an active mechanism for emotion induction. Furthermore, for the success and quality of research settings, VR must select the appropriate material to effectively evoke emotions. Therefore, in the present review, we evaluated to what extent VR visual and audio-visual stimuli, games, and tasks, and 360-degree panoramas and videos can elicit emotions based on the current literature. Further, we present public datasets generated by VR and emotion-sensing interfaces that can be used in VR based research. The conclusions of this survey reveal that VR has a great potential to evoke emotions effectively and naturally by generating motivational and empathy mechanisms which makes it an ecologically valid paradigm to study emotions.

cs.HC↗