arXiv subjects

Michel Dumontier

Publications and source records attributed to Michel Dumontier.

At least 19 recordsLinked to original sources

arXiv2026

CompDiff enables fair and zero shot medical image generation across demographic intersections through compositional diffusion

Medical image generators trained on imbalanced data can fail at demographic intersections absent from training. We introduce CompDiff, which encodes age, sex and race separately and composes supervised demographic tokens alongside clinical text. Across chest radiographs and fundus images, CompDiff improves overall and subgroup fidelity relative to prompt conditioning (RoentGen-v2) and loss reweighting (FairDiffusion). It generalises in zero-shot generation to 16 chest X-ray intersections excluded from training, achieving the lowest mean FID-RadImageNet in every intersection. In a blinded reader study of these unseen intersections, two radiologists gave CompDiff the highest mean scores among generators for anatomical realism and agreement with the clinical impression, and selected its images most often as the most realistic. Pretraining with CompDiff images improved downstream classification, while CompDiff audit cohorts reduced estimation error on rare intersections. These findings support compositional demographic conditioning for extending medical image synthesis to underserved populations. Code: https://github.com/mahmoudibrahim98/CompDiff

cs.CV↗

arXiv2026

Demographically-Conditioned Synthetic Medical Images for Bias Mitigation and Bias Detection in Disease Classifiers

Per-subgroup fairness audits of medical image classifiers face a sample-size problem: minority subgroups in held-out test sets have so few samples that the resulting confidence intervals on per-subgroup performance are wider than the bias the audit is meant to detect. We argue that a demographically-conditioned synthetic generator can do both: mitigate bias on the training side and detect bias on the evaluation side. Working on COVID-19 chest CT classification with an end-to-end fine-tuned Stable Diffusion 2.1 generator, we make two findings. For bias mitigation (training), a demographically-balanced synthetic cohort is most useful as a pretraining prior, not as joint augmentation: with the same fixed data, sequential pretraining followed by fine-tuning substantially outperforms joint augmentation, and the resulting classifier surpasses the full-real baseline at $\sim$$100\times$ real-data efficiency. For bias detection (evaluation), across five synthetic minority cohorts and five classifier seeds, the synthetic estimator reproduces the subgroup ranking of a well-powered real oracle (Spearman $ρ= 1.00$ on MCC and Recall) and gives the more reliable per-cell estimate where the small real test set runs out of samples. The synthetic cohort is therefore most useful in exactly the cells that fairness audits care about, as both a fix for and a measure of subgroup bias.

cs.AI↗

arXiv2026

LLM-Guided Monte Carlo Tree Search over Knowledge Graphs: Composing Mechanistic Explanations for Drug-Disease Pairs

Extracting multi-step explanations from knowledge graphs poses a combinatorial challenge requiring both heuristic guidance (as candidates proliferate with depth) and credit assignment (as path quality emerges over extended sequences). Frontier LLMs, strong on knowledge/reasoning benchmarks, offer a compelling source of such heuristics, yet their knowledge comes sans guarantees and compositional performance degrades as chains lengthen. We thus present TESSERA, a 3-part neuro-symbolic framework that uses LLMs in a circumscribed role: for local discriminative judgement rather than autonomous multi-step generation; the knowledge graph then defines the hypothesis space enforcing hard structural constraints, and MCTS coordinates the long-horizon search with principled credit assignment via backpropagation. LLMs perform dual roles as a prior policy biasing exploration and a comparative state evaluator supplying reward signals. Evaluation on drug mechanism elucidation across two complementary knowledge graphs demonstrates fidelity to curated biology while surfacing coherent alternative mechanisms, with ablations confirming discriminative contribution from both LLM components. Beyond its current application, our framework offers a general paradigm for compositional reasoning over structured knowledge.

cs.AI↗

arXiv2026

AudioToolAgent: An Agentic Framework for Audio-Language Models

Large Audio-Language Models (LALMs) perform well on audio understanding tasks but lack multistep reasoning and tool-calling found in recent Large Language Models (LLMs). This paper presents AudioToolAgent, a framework that coordinates audio-language models as tools via a central LLM agent that accesses tool adapters for audio question answering and speech-to-text. The agent reasons about which tools to invoke, how to formulate follow-up queries, and how to arbitrate conflicting tool outputs, without accessing the audio. Experiments with MMAU, MMAR, and MMAU-Pro show state-of-the-art accuracy: up to 77.50% in MMAU, 77.00% in MMAR, and 61.90% in MMAU-Pro. Shapley-based analysis identifies effective agent-tool combinations. The code and reproduction materials are available at https://github.com/GLJS/AudioToolAgent.

cs.SD↗

arXiv2025

Enabling Granular Subgroup Level Model Evaluations by Generating Synthetic Medical Time Series

We present a novel framework for leveraging synthetic ICU time-series data not only to train but also to rigorously and trustworthily evaluate predictive models, both at the population level and within fine-grained demographic subgroups. Building on prior diffusion and VAE-based generators (TimeDiff, HealthGen, TimeAutoDiff), we introduce \textit{Enhanced TimeAutoDiff}, which augments the latent diffusion objective with distribution-alignment penalties. We extensively benchmark all models on MIMIC-III and eICU, on 24-hour mortality and binary length-of-stay tasks. Our results show that Enhanced TimeAutoDiff reduces the gap between real-on-synthetic and real-on-real evaluation (``TRTS gap'') by over 70\%, achieving $Δ_{TRTS} \leq 0.014$ AUROC, while preserving training utility ($Δ_{TSTR} \approx 0.01$). Crucially, for 32 intersectional subgroups, large synthetic cohorts cut subgroup-level AUROC estimation error by up to 50\% relative to small real test sets, and outperform them in 72--84\% of subgroups. This work provides a practical, privacy-preserving roadmap for trustworthy, granular model evaluation in critical care, enabling robust and reliable performance analysis across diverse patient populations without exposing sensitive EHR data, contributing to the overall trustworthiness of Medical AI.

cs.LG↗

arXiv2025

AudSemThinker: Enhancing Audio-Language Models through Reasoning over Semantics of Sound

Audio-language models have shown promising results in various sound understanding tasks, yet they remain limited in their ability to reason over the fine-grained semantics of sound. In this paper, we present AudSemThinker, a model whose reasoning is structured around a framework of auditory semantics inspired by human cognition. To support this, we introduce AudSem, a novel dataset specifically curated for semantic descriptor reasoning in audio-language models. AudSem addresses the persistent challenge of data contamination in zero-shot evaluations by providing a carefully filtered collection of audio samples paired with captions generated through a robust multi-stage pipeline. Our experiments demonstrate that AudSemThinker outperforms state-of-the-art models across multiple training settings, highlighting its strength in semantic audio reasoning. Both AudSemThinker and the AudSem dataset are released publicly.

cs.SD↗

arXiv2025

Data-Balanced Curriculum Learning for Audio Question Answering

Audio question answering (AQA) requires models to understand acoustic content and perform complex reasoning. Current models struggle with dataset imbalances and unstable training dynamics. This work combines curriculum learning with statistical data balancing to address these challenges. The method labels question difficulty using language models, then trains progressively from easy to hard examples. Statistical filtering removes overrepresented audio categories, and guided decoding constrains outputs to valid multiple-choice formats. Experiments on the DCASE 2025 training set and five additional public datasets show that data curation improves accuracy by 11.7% over baseline models, achieving 64.2% on the DCASE 2025 benchmark.

cs.SD↗

arXiv2025

Adapting Lightweight Vision Language Models for Radiological Visual Question Answering

Recent advancements in vision-language systems have improved the accuracy of Radiological Visual Question Answering (VQA) Models. However, some challenges remain across each stage of model development: limited expert-labeled images hinders data procurement at scale; the intricate and nuanced patterns of radiological images make modeling inherently difficult; and the lack of evaluation evaluation efforts makes it difficult to identify cases where the model might be ill-conditioned. In this study, we fine-tune a lightweight 3B parameter vision-language model for Radiological VQA, demonstrating that small models, when appropriately tuned with curated data, can achieve robust performance across both open- and closed-ended questions. We propose a cost-effective training pipeline from synthetic question-answer pair generation to multi-stage fine-tuning on specialised radiological domain-targeted datasets (e.g., ROCO v2.0, MedPix v2.0). Our results show that despite operating at a fraction of the scale of state-of-the-art models such as LLaVA-Med, our model achieves promising performance given its small parameter size and the limited scale of training data. We introduce a lightweight saliency-based diagnostic tool that enables domain experts to inspect VQA model performance and identify ill-conditioned failure modes through saliency analysis.

cs.CV↗

arXiv2025

CDE-Mapper: Using Retrieval-Augmented Language Models for Linking Clinical Data Elements to Controlled Vocabularies

The standardization of clinical data elements (CDEs) aims to ensure consistent and comprehensive patient information across various healthcare systems. Existing methods often falter when standardizing CDEs of varying representation and complex structure, impeding data integration and interoperability in clinical research. We introduce CDE-Mapper, an innovative framework that leverages Retrieval-Augmented Generation approach combined with Large Language Models to automate the linking of CDEs to controlled vocabularies. Our modular approach features query decomposition to manage varying levels of CDEs complexity, integrates expert-defined rules within prompt engineering, and employs in-context learning alongside multiple retriever components to resolve terminological ambiguities. In addition, we propose a knowledge reservoir validated by a human-in-loop approach, achieving accurate concept linking for future applications while minimizing computational costs. For four diverse datasets, CDE-Mapper achieved an average of 7.2\% higher accuracy improvement compared to baseline methods. This work highlights the potential of advanced language models in improving data harmonization and significantly advancing capabilities in clinical decision support systems and research.

cs.IR↗

arXiv2025

Semantic Web and Creative AI -- A Technical Report from ISWS 2023

The International Semantic Web Research School (ISWS) is a week-long intensive program designed to immerse participants in the field. This document reports a collaborative effort performed by ten teams of students, each guided by a senior researcher as their mentor, attending ISWS 2023. Each team provided a different perspective to the topic of creative AI, substantiated by a set of research questions as the main subject of their investigation. The 2023 edition of ISWS focuses on the intersection of Semantic Web technologies and Creative AI. ISWS 2023 explored various intersections between Semantic Web technologies and creative AI. A key area of focus was the potential of LLMs as support tools for knowledge engineering. Participants also delved into the multifaceted applications of LLMs, including legal aspects of creative content production, humans in the loop, decentralised approaches to multimodal generative AI models, nanopublications and AI for personal scientific knowledge graphs, commonsense knowledge in automatic story and narrative completion, generative AI for art critique, prompt engineering, automatic music composition, commonsense prototyping and conceptual blending, and elicitation of tacit knowledge. As Large Language Models and semantic technologies continue to evolve, new exciting prospects are emerging: a future where the boundaries between creative expression and factual knowledge become increasingly permeable and porous, leading to a world of knowledge that is both informative and inspiring.

cs.AI↗

arXiv2025

Audio-Language Datasets of Scenes and Events: A Survey

Audio-language models (ALMs) generate linguistic descriptions of sound-producing events and scenes. Advances in dataset creation and computational power have led to significant progress in this domain. This paper surveys 69 datasets used to train ALMs, covering research up to September 2024 (https://github.com/GLJS/audio-datasets). It provides a comprehensive analysis of datasets origins, audio and linguistic characteristics, and use cases. Key sources include YouTube-based datasets like AudioSet with over two million samples, and community platforms like Freesound with over 1 million samples. Through principal component analysis of audio and text embeddings, the survey evaluates the acoustic and linguistic variability across datasets. It also analyzes data leakage through CLAP embeddings, and examines sound category distributions to identify imbalances. Finally, the survey identifies key challenges in developing large, diverse datasets to enhance ALM performance, including dataset overlap, biases, accessibility barriers, and the predominance of English-language content, while highlighting opportunities for improvement.

cs.SD↗

arXiv2024

Generative AI for Synthetic Data Across Multiple Medical Modalities: A Systematic Review of Recent Developments and Challenges

This paper presents a comprehensive systematic review of generative models (GANs, VAEs, DMs, and LLMs) used to synthesize various medical data types, including imaging (dermoscopic, mammographic, ultrasound, CT, MRI, and X-ray), text, time-series, and tabular data (EHR). Unlike previous narrowly focused reviews, our study encompasses a broad array of medical data modalities and explores various generative models. Our search strategy queries databases such as Scopus, PubMed, and ArXiv, focusing on recent works from January 2021 to November 2023, excluding reviews and perspectives. This period emphasizes recent advancements beyond GANs, which have been extensively covered previously. The survey reveals insights from three key aspects: (1) Synthesis applications and purpose of synthesis, (2) generation techniques, and (3) evaluation methods. It highlights clinically valid synthesis applications, demonstrating the potential of synthetic data to tackle diverse clinical requirements. While conditional models incorporating class labels, segmentation masks and image translations are prevalent, there is a gap in utilizing prior clinical knowledge and patient-specific context, suggesting a need for more personalized synthesis approaches and emphasizing the importance of tailoring generative approaches to the unique characteristics of medical data. Additionally, there is a significant gap in using synthetic data beyond augmentation, such as for validation and evaluation of downstream medical AI models. The survey uncovers that the lack of standardized evaluation methodologies tailored to medical images is a barrier to clinical application, underscoring the need for in-depth evaluation approaches, benchmarking, and comparative studies to promote openness and collaboration.

cs.LG↗

arXiv2024

Improving rule mining via embedding-based link prediction

Rule mining on knowledge graphs allows for explainable link prediction. Contrarily, embedding-based methods for link prediction are well known for their generalization capabilities, but their predictions are not interpretable. Several approaches combining the two families have been proposed in recent years. The majority of the resulting hybrid approaches are usually trained within a unified learning framework, which often leads to convergence issues due to the complexity of the learning task. In this work, we propose a new way to combine the two families of approaches. Specifically, we enrich a given knowledge graph by means of its pre-trained entity and relation embeddings before applying rule mining systems on the enriched knowledge graph. To validate our approach, we conduct extensive experiments on seven benchmark datasets. An analysis of the results generated by our approach suggests that we discover new valuable rules on the enriched graphs. We provide an open source implementation of our approach as well as pretrained models and datasets at https://github.com/Jean-KOUAGOU/EnhancedRuleLearning

cs.AI↗

arXiv2024

ACES: Evaluating Automated Audio Captioning Models on the Semantics of Sounds

Automated Audio Captioning is a multimodal task that aims to convert audio content into natural language. The assessment of audio captioning systems is typically based on quantitative metrics applied to text data. Previous studies have employed metrics derived from machine translation and image captioning to evaluate the quality of generated audio captions. Drawing inspiration from auditory cognitive neuroscience research, we introduce a novel metric approach -- Audio Captioning Evaluation on Semantics of Sound (ACES). ACES takes into account how human listeners parse semantic information from sounds, providing a novel and comprehensive evaluation perspective for automated audio captioning systems. ACES combines semantic similarities and semantic entity labeling. ACES outperforms similar automated audio captioning metrics on the Clotho-Eval FENSE benchmark in two evaluation categories.

cs.SD↗

arXiv2023

Models towards Risk Behavior Prediction and Analysis: A Netherlands Case study

In many countries financial service providers have to elicit their customers risk preferences, when offering products and services. For instance, in the Netherlands pension funds will be legally obliged to factor in their clients risk preferences when devising their investment strategies. Therefore, assessing and measuring the risk preferences of individuals is critical for the analysis of individuals' behavior and policy prescriptions. In the psychology and economics, a number of methods to elicit risk preferences have been developed using hypothetical scenarios and economic experiments. These methods of eliciting individual risk preferences are usually applied to small samples because they are expensive and the implementation can be complex and not suitable when large cohorts need to be measured. A large number of supervised learning models ranging from linear regression to support vector machines are used to predict risk preference measures using socio-economic register data such as age, gender, migration background and other demographic variables in combination with data on income, wealth, pension fund contributions, and other financial data. The employed machine learning models cover a range of assumptions and properties as well as a diverse set of regression metrics. The optimum model is selected using the metrics and interpretability of the model. The optimal models are lasso regression and gradient boosting machines with mean average percentage error of about 30%. This is important as it helps to estimate risk attitudes without actually measuring them. It should be noted that with the current accuracy the tested models are not ready for deployment for applications that require high accuracy. However, the results do indicate which models should be used in situations that do not require the most accurate predictions such as augmentation data for pensions' recommendation.

cs.CE↗

arXiv2023

FAIR Begins at home: Implementing FAIR via the Community Data Driven Insights

Arguments for the FAIR principles have mostly been based on appeals to values. However, the work of onboarding diverse researchers to make efficient and effective implementations of FAIR requires different appeals. In our recent effort to transform the institution into a FAIR University by 2025, here we report on the experiences of the Community of Data Driven Insights (CDDI). We describe these experiences from the perspectives of a data steward in social sciences and a data scientist, both of whom have been working in parallel to provide research data management and data science support to different research groups. We initially identified 5 challenges for FAIR implementation. These perspectives show the complex dimensions of FAIR implementation to researchers across disciplines in a single university.

cs.DL↗

arXiv2023

TAPS Responsibility Matrix: A tool for responsible data science by design

Data science is an interdisciplinary research area where scientists are typically working with data coming from different fields. When using and analyzing data, the scientists implicitly agree to follow standards, procedures, and rules set in these fields. However, guidance on the responsibilities of the data scientists and the other involved actors in a data science project is typically missing. While literature shows that novel frameworks and tools are being proposed in support of open-science, data reuse, and research data management, there are currently no frameworks that can fully express responsibilities of a data science project. In this paper, we describe the Transparency, Accountability, Privacy, and Societal Responsibility Matrix (TAPS-RM) as framework to explore social, legal, and ethical aspects of data science projects. TAPS-RM acts as a tool to provide users with a holistic view of their project beyond key outcomes and clarifies the responsibilities of actors. We map the developed model of TAPS-RM with well-known initiatives for open data (such as FACT, FAIR and Datasheets for datasets). We conclude that TAPS-RM is a tool to reflect on responsibilities at a data science project level and can be used to advance responsible data science by design.

cs.CY↗

arXiv2022

Improving Correlation Capture in Generating Imbalanced Data using Differentially Private Conditional GANs

Despite the remarkable success of Generative Adversarial Networks (GANs) on text, images, and videos, generating high-quality tabular data is still under development owing to some unique challenges such as capturing dependencies in imbalanced data, optimizing the quality of synthetic patient data while preserving privacy. In this paper, we propose DP-CGANS, a differentially private conditional GAN framework consisting of data transformation, sampling, conditioning, and networks training to generate realistic and privacy-preserving tabular data. DP-CGANS distinguishes categorical and continuous variables and transforms them to latent space separately. Then, we structure a conditional vector as an additional input to not only presents the minority class in the imbalanced data, but also capture the dependency between variables. We inject statistical noise to the gradients in the networking training process of DP-CGANS to provide a differential privacy guarantee. We extensively evaluate our model with state-of-the-art generative models on three public datasets and two real-world personal health datasets in terms of statistical similarity, machine learning performance, and privacy measurement. We demonstrate that our model outperforms other comparable models, especially in capturing dependency between variables. Finally, we present the balance between data utility and privacy in synthetic data generation considering the different data structure and characteristics of real-world datasets such as imbalance variables, abnormal distributions, and sparsity of data.

cs.LG↗