Funding
Competitive grants and mandates supporting my research as a principal investigator. Each title links to the grant’s official page where one exists. The full record is on ORCID. Below the divider are earlier grants I contributed to without leading, stating the role I played.
IMAGIN-AIR: From Images to Knowledge: AI to Elucidate Mechanisms and Improve Therapeutic Strategies for Retinal Inflammation
About this grant
Context: Uveitis, an inflammatory disease of the eye that remains poorly understood, primarily affects young adults and leads to major socioeconomic consequences such as high medical costs, loss of productivity, and reduced quality of life. It accounts for 5–20% of cases of legal blindness in Europe and the United States, predominantly affecting a young population and causing lasting impacts on their autonomy and quality of life. Fluorescein angiography (FA) is the main diagnostic tool for evaluating this inflammation. However, its interpretation remains complex, subjective, and time-consuming, limiting its use in clinical trials and routine practice despite its critical role in therapeutic decision-making. Uveitis may have autoimmune, infectious, traumatic, or idiopathic origins, further complicating its clinical management. Overall objective: The IMAGIN-AIR project aims to develop an artificial intelligence (AI)–based medical device capable of automatically analyzing FA examinations to assist clinicians in the diagnosis and complex management of uveitis. The AI will enable objective grading of retinal inflammation, improved assessment of disease progression, and support for personalized treatment strategies. A PhD student will be specifically integrated into the project to ensure the crucial link between technical development and clinical application. Thanks to CARIGEST funding, the student will receive advanced training in AI applied to medical imaging and interdisciplinary collaboration, strengthening her skills and expertise for a future career in this innovative field.
AI2Pub: AI Insights to the Public through bi-directional media channels
About this grant
Total awarded: CHF 200,000
Artificial intelligence (AI) has become a powerful and pervasive technology, influencing many aspects of our daily lives. However, its rapid growth and integration into society raise complex questions and concerns. Our team of scientists and communication experts are committed to improving public understanding of AI technologies, fostering a positive societal impact.Building on the foundations of our previous project, NewsOnAI, we will go beyond traditional communication channels such as workshops and media publications. Our proactive engagement efforts will establish dynamic, two-way communication between scientists and the public, utilizing new formats like interactive exhibitions and theater performances.Analyzing feedback from these initiatives will enable scientists to pursue research directions that effectively address societal concerns. Over time, we anticipate that our efforts will create a multiplying social impact, fostering informed public discourse and a deeper understanding of AI technologies.
FairMI: Machine Learning Fairness with Application to Medical Images
About this grant
Total awarded: CHF 577,855
The utilisation of Artificial Intelligence (AI) in healthcare is exponentially growing, particularly in the field of medical imaging. AI involves training computer systems to recognise and analyse patterns. However, a paramount concern arises: the potential bias of these ML algorithms towards specific demographic groups. A situation where a machine is more proficient in diagnosing a condition in one demographic but fails in another due to inherent biases is not only ethically problematic but also poses a significant threat to patient care. The FairMI project is committed to establishing standards and protocols, ensuring that the integration of AI tools in healthcare is both safe and equitable for all patients. This endeavor is anticipated to pave the way for the conscientious and responsible employment of artificial intelligence in medical care and beyond.
DREAM: Detection and Research for Learning Sets on Electroencephalographic Patterns
About this grant
Total awarded: CHF 99,897
Project DREAM aims to develop a scalable foundation model for multi-channel EEG signal analysis, spanning from basic polysomnography to high-density and intracranial recordings. The goal is to create a unified AI model capable of adapting to various EEG configurations while maintaining simplicity, interpretability, and computational efficiency. By leveraging convolutional neural networks and attention mechanisms, the model will extract robust features for diverse tasks, including sleep stage classification, epilepsy detection, and cognitive function assessment. Collaborations with leading hospitals will provide real-world datasets for validation, ensuring the model’s generalization. Ultimately, this project seeks to advance AI-driven biomedical signal processing, enhancing diagnostic and monitoring capabilities in neurology and sleep medicine.
VALIDATE-H: VALaIs aDaptation of Artificial inTelligence modEls in Healthcare
About this grant
Total awarded: CHF 115,980
Can we deploy artificial intelligence (AI) models developed in other countries, e.g. China, in the Valais healthcare system? Through the Cantonal Prevention and Health Promotion Strategy, Horizon 2030, the canton of Valais aims to enhance the effectiveness of prevention and health promotion for the entire population. The initial intersectional focus in PPS2030 addresses the principle of equal opportunities, highlighting the urgent need to include all demographic groups—especially vulnerable populations—in health-related initiatives. Project VALIDATE-H will develop an evaluation framework for artificial intelligence models in healthcare (AIMH) with two main objectives: 1. Validate and monitor the clinical utility of AIMH, pre-trained with international datasets, in population-specific settings and applications relevant for the Valais healthcare system, and 2. Fine-tune these models to mitigate potential demographic biases in a fair, transparent and accountable way. The project will first select target population groups and applications for AIMH, then, using distribution metrics, fairness criteria and explainability, VALIDATE-H will match local settings to those found on international datasets and models. VALIDATE-H will allow for a continuous assessment of AIMH, monitoring and updating the knowledge regarding potential barriers that might hinder the adaptation and adoption of artificial intelligence in the Valaisan healthcare system. Our consortium is well-equipped to tackle the objectives mentioned above with complementary expertise on machine learning, computer vision and medical data processing (Idiap), together with distributed and explainable AI, and multi-agent systems (HES-SO). Finally, the expertise from our partner in the Observatoire Valaisan de la Santé (OVS), will ground the direction of the project into concrete applications for the Valais population and cantonal health systems.
CollabCloud: A Collaborative Research Cloud Infrastructure powering Discovery and Exchange
About this grant
Total awarded: CHF 606,229
Collaborative breakthrough research in Artificial Intelligence (AI) requires access to well-tuned systems, specific computing, large datasets, and dedicated storage, often only accessible to a select few. This reality slows down complementary research, as a substantial amount of project time is often dedicated to repeating setups, understanding computation and storage intricacies, data properties, and how to properly transfer systems between institutions, to pursue overall objectives. Examples can be often found in cases encompassing consortia with computational versus non-computational specialists, or containing a virtuous mixture of domain experts (medical doctors, biologists, psychologists) and data scientists. Reproducibility and technology transfer are essential tools in thriving projects, however these concepts are costly to implement and maintain. Deploying AI solutions requires multidisciplinary expertise, the right hardware, and AI specialists to tune and ready tools for collaborative use. In practice, lack of specific expertise slowdowns partner-to-partner communication affecting overall productivity. In projects with industry or government-academia partnerships driven by concrete societal needs, replicating project conclusions with different, private datasets, or allowing partners to infer from pre-trained models using adequate hardware setups, is often avoided because of these barriers. Moreover, the growing scale, complexity and impact of contemporary AI systems such as Large Language Models (LLMs) accelerates the need for accessible infrastructures which can guarantee systematised, transparent and increased collaborative work. We intend to bridge these gaps by building “CollabCloud", a cloud-based research infrastructure to boost collaborative research for current and future projects at the Idiap Research Institute. The main focal points will be boosting the ability of easily exporting research workflows, exploring AI models by both computational and non-computational experts, and allowing controlled access to shared storage and computing power for collaborative projects. Beyond these goals, CollabCloud will enable Idiap to participate in developing important topics shaping the future of AI, such as Federated Learning, and cloud-based scientfic networks, which require connectivity and storage capabilities adapted to such purposes. As discussed in the institutional support letter, this vision aligns well with Idiap’s future, its predicted growth (with a current data center reaching the limits of its maximum capacity), and the notion of Cross Research Groups (CRG), that is part of our 2021-2024 Research Program as approved by the Federal State Secretariat for Education, Research, and Innovation (SERI).
NewsOnAI: Scientific voice to latest news of AI technology
About this grant
Total awarded: CHF 50,000
We are a group of scientists who joined forces to enable artificial intelligence (AI) technologies to make a greater impact on Swiss society by participating in a public debate over the latest news. We plan to publish research-based opinion articles on mainstream media in Switzerland, collect the readers’ feedback on the published articles, and follow up by organizing live seminars for the audience in the area. We will also post rounded responses on social media to ensure bi-directional dialogue. Our primary target audience is Swiss citizens reading a newspaper. They are mostly non-scientific public and may not have specific knowledge of AI technologies. We often receive news about new products, companies, and social issues related to AI technologies. An example is a controversy in the U.K. over an algorithm used to substitute university-entrance exams in 2020. Another example is an accident in testing an autonomous driving car. Usually, the public’s perception of information is affected by the tones and perspectives taken by the writers. News reporters are not necessarily technical experts and are often incentivized to publish striking news stories to attract more audiences. As some entrepreneurs, celebrity CEOs, and politicians may try to magnify the public hype for their benefit, the conflict of interests can influence media perspectives. Therefore, there is a need for AI scientists to collaborate with journalists to provide evidence-based analysis in public debate whenever appropriate. Without such cooperation, the general audience would miss out on essential information regarding AI. Scientists can help the public better understand AI technological issues by clarifying the potential capabilities and limitations.Not many people would read scientists’ research published in academic journals. However, publishing news articles has a greater impact, as the article is distributed to a much broader audience in Switzerland and worldwide through popular media platforms. In addition, follow-up activities at a local level will ensure the interactive and bi-directional dialogue between scientists and the readers. The analysis of the feedback will allow the project team to understand the public’s perspectives in facing future technologies. In turn, the acquired understanding can guide scientists to pursue an impactful research direction. After this Agora project, the current members will be able to form a new team to engage other scientists to study societal trends, raise their voices, and provide research-based analysis to guide the public about future technologies. The social impact will be multiplied over time.
Nouvelle plateforme de recherche « Cloud » pour étendre les capacités de recherche, d’échange et de formation de l’Idiap
About this grant
Total awarded: CHF 285,000
L’Institut de Recherche Idiap est l’un de précurseurs dans le domaine de l’Intelligence Artificielle (IA). Ce domaine fait cependant face à de nouveaux challenges avec des besoins en ressources de calcul de plus en plus extrêmes. À ce fait, se rajoutent deux nouvelles caractéristiques : les collaborations se font de plus en plus pluridisciplinaires, comptant des acteurs avec peu, voire aucune expertise liée à l’informatique, et une croissante demande d’ouverture et reproductibilité, de la part de la communauté scientifique. Afin de dynamiser ses futurs projets, l’institut souhaite construire une infrastructure de recherche collaborative innovante « CollabCloud », basée sur « l’informatique dans les nuages », pour laquelle nous demandons votre soutien. Pour démontrer le potentiel de cette plateforme, nous proposons aussi un sous projet de démonstration de l’applicabilité de ce système pour la robotique et plus particulièrement l'interaction humain-robot.
CAD4IED: Computer Aided Diagnosis for Inflammatory Eye Diseases
About this grant
Total awarded: CHF 200,000
Fluorescein angiography (FA) is a unique tool for the analysis of the retinal vasculature in both its morphology and its function: it is the only clinical method that allows to evaluate the function and integrity of the blood-retinal barrier, providing means for the detection and grading of inflammatory diseases affecting blood vessels in the eye (Vasculitis). FA is an invasive technique involving a non-negligible level of risk for the patient. The interpretation of an angiography is intrinsically challenging, requiring years of clinical experience. The overarching objective of this proposal is to develop and (clinically) validate a system to automatically detect and grade inflammatory eye diseases, with minimal risk, allowing for better patient management and care. This proposal also represents the first attempt to automatically grade angiography images and holds the potential to help doctors in the challenging interpretation of these data, delivering cutting edge machine learning solutions that would support clinicians at our hospitals, and benefit the scientific community at large. To reduce risk for the patient, we will search for novel biomarkers from other ophthalmic imaging modalities, such as fundus images, that would be less invasive, cheaper and safer to acquire, but correlate well with inflammatory diseases. In this pilot project, we propose to establish data, annotations, and develop a prototype grading system (machine learning model), to automatically evaluate inflammatory signs from FA images. A preliminary study on novel biomarkers from alternate modalities will complement this phase of our project. We intend to disseminate our work via scientific articles to be submitted to medical and computational journals. If this pilot is successful, a follow-up project will be submitted.
LEARN-REAL: LEARNing physical manipulation skills with simulators using REAListic variations
About this grant
Total awarded: CHF 301,936
The acquisition of manipulation skills in robotics involves the combination of object recognition, action-perception coupling and physical interaction with the environment. Several learning strategies have been proposed to acquire such skills. As for humans and other animals, the robot learner needs to be exposed to varied situations. It needs to try and refine the skill many times, and/or needs to observe several attempts of successful movements by others to adapt and generalize the learned skill to new situations. Such skill is not acquired in a single training cycle, motivating the need to compare, share and re-use the experiments. In LEARN-REAL, we propose to learn manipulation skills through simulation for object, environment and robot, with an innovative toolset comprising: 1) a simulator with realistic rendering of variations allowing the creation of datasets and the evaluation of algorithms in varied situations; 2) a virtual-reality interface to interact with the robots within a controlled virtual environment, to teach robots object manipulation skills in multiple configurations; and 3) a web-based infrastructure for principled, reproducible and transparent benchmarking of learning algorithms for object recognition and manipulation by robots. These features will extend existing softwares in several ways. 1) and 2) will capitalize on the widespread development of realistic simulators for the gaming industry and the associated low-cost virtual reality interfaces. 3) will harness the existing BEAT toolchain developed at Idiap, which will be extended to object recognition and manipulation by robots, including the handling of data, algorithms and benchmarking results. As use case, we will study the scenario of vegetable/fruit picking and sorting.
SECure: Safe and Explainable Clinical AI for Orthopaedic Surgical Assessment
About this grant
The main aim of SECure is to design a safe and explainable Machine Learning-based system that supports a qualified physician to decide whether or not a orthopaedic surgery is indicated for a given patient.
SEWS2: Smart Early Warning Score System for in- and out-hospital care via anomaly detection
About this grant
Total awarded: CHF 15,000
The company Vtuls propose to use a device (Vtuls Biosensor) and software-powered machine learning models ("virtual nurse") to continuously monitor Vital signs in patients to provide Early Warning alerts of potential patient health deterioration. The current hospital alert systems are based on rule-based Algorithms applied independently on a single Vital sign (Blood pressure, Body temp, Respiration rate, Sp02, Heart Rate) and also only on the current spot measurement of that Vital sign. We believe applying AI/Machine learning models on the combination of the Multiple Vital signs taken together and with their history will lead to an improvement in time and accuracy of early detection of potential health deterioration. The following use-cases are contemplated in this project: 1) Assist nurses in hospital settings to detect early signs of medical condition deterioration in patients ; 2) Evaluate and potentially reduce patient discharging times ; 3) Continuously monitor patients after discharge for deteriorating conditions (early warning)
Earlier career contributions
Mostly while working at the Biometrics & Security group of Sébastien Marcel.
AI4EU: A European AI On Demand Platform and Ecosystem
About this grant
Roles: Co-authorship of the Idiap contribution, project management
Artificial Intelligence is a disruptive technology of our times with expected impacts rivalling those of electricity or printing. Resources for innovation are currently dominated by giant tech companies in North America and China. To ensure European independence and leadership, we must invest wisely by bundling, connecting and opening our AI resources. AI4EU will efficiently build a comprehensive European AI-on-demand platform to lower barriers to innovation, to boost technology transfer and catalyse the growth of start-ups and SMEs in all sectors through Open calls and other actions. The platform will act as a broker, developer and one-stop shop providing and showcasing services, expertise, algorithms, software frameworks, development tools, components, modules, data, computing resources, prototyping functions and access to funding. Training will enable different user communities (engineers, civic leaders, etc.) to obtain skills and certifications. The AI4EU Platform will establish a world reference, built upon and interoperable with existing AI and data components (e.g. the Acumos open-source framework, QWT search engine..) and platforms. It will mobilize the whole European AI ecosystem and already unites 80 partners in 21 countries including researchers, innovators and related talents. Eight industry-driven AI pilots will demonstrate the value of the platform as an innovation tool. In order to enhance the platform, research on five key interconnected AI scientific areas will be carried out using platform technologies and results will be implemented. The pilots and research will showcase how AI4EU can stimulate scientific discovery and technological innovation. The AI4EU Ethical Observatory will be established to ensure the respect of human centred AI values and European regulations. Sustainability will be ensured via the creation of the AI4EU Foundation. The results will feed a new and comprehensive Strategic Research Innovation Agenda for Europe.
ALLIES: Autonomous Lifelong learnIng intelLigent Systems
About this grant
Roles: Co-authorship, project management, implementation
Total awarded: CHF 496,620
Standard machine learning systems require massive data and huge processing infrastructures, but the main limitation to their spreading comes from the need of the empirical and rare knowledge of an experienced data scientist able to set and adjust their behavior over time. The ALLIES project will lay the foundation for development of autonomous intelligent systems sustaining their performance across time. Such unsupervised system will be able to auto-update and perform self-evaluation to be aware of the evolution of its own knowledge acquisition. It should adapt to a changing environment by following a given learning scenario that balances the importance of performance on past and present data to avoid unwanted regression. Such systems could not be developed without adapted metrics and protocols enabling their objective and reproducible evaluation. This evaluation should assess the performance on the given task and quantify the effort required to reach it in terms of unsupervised data collected by the system and of interaction with humans in the case of active-learning. The ALLIES project will develop and disseminate those metrics and protocols. They will be available to european actors via an open evaluation platform dedicated to reproducible research. An evaluation campaign and a workshop will be organised to engage the community on this path. By publicly releasing the evaluation protocols and data, by releasing a dedicated evaluation platform and by developing autonomous systems for two tasks: machine translation and speaker diarization, we believe that the ALLIES project will boost the development of intelligent lifelong learning systems in Europe.
BATL: Biometric Authentication with a Timeless Learner
About this grant
Roles: Co-authorship of the Idiap participation (face modality)
BATL built an end-to-end multimodal presentation attack detection system for IARPA's Odin program, pairing face, iris and fingerprint sensors with machine learning so that both known and unknown attacks on biometric recognition could be detected. The originality of the proposal was a multi-channel acquisition system, which hardened the spoofable surface of face recognition far beyond what a single visible-light camera exposes. The consortium was led by USC's Information Sciences Institute; Idiap was responsible for the face modality, carrying over the lessons learned during TABULA RASA. The project was funded in full to its end, which demonstrated those lessons were sound.
BEAT: Biometrics Evaluation and Testing
About this grant
Roles: Co-authorship, implementation, project management
Identity management using Biometrics is deployed in a growing number of applications ranging from identification platforms (e.g. biometric passports) to access control systems for border checks or banking transactions. Unfortunately, the reliability of these technologies remains difficult to compare. There are no european-wide standards for evaluating their accuracy, their robustness to attacks or their privacy preservation strength. BEAT will fill this gap by building an online and open platform to transparently evaluate biometric systems, designing protocols and tools for vulnerability analysis and developing standardization documents for Common Criteria evaluations. The planned impact is three-fold: the reliability of biometric systems becomes standardized and comparable, potentially leading to a meaningful increase in their performance; technology transfer from research to companies becomes easier with the use of an interoperable framework; authorities and decision-makers become more informed about the progress made in biometrics as results impact standards.
COHFACE: COntactless Heartbeat detection for trustworthy FACE Biometrics
About this grant
Roles: Co-authorship, dataset acquisition management, implementation
Total awarded: CHF 50,000
Identity spoofing is a contender for high-security face recognition applications. Nowadays, our images are wide-spread on the internet and can be potentially used to attack biometric systems. The latest developments in counter-measures for facial spoofing fail to recognize attacks in challenging conditions. Recent advances in the field of contactless cardiac heart rate measurement using commodity hardware opens up new possibilities for face biometrics anti-spoofing. Unfortunately, current research in commodity photoplethysmography (PPG) is not reproducible. Deployment for face recognition also requires addressing speed and illumination robustness. We propose to conduct a reproducible research study with a new public database to be collected, including user videos and related biomedical signals. We will develop baseline systems for the detection of heart rates based on current state of the art. We will leverage from in-house know-how in face detection, and tracking to design an open-source processing chain for commodity PPG applied to face biometrics anti-spoofing that can be used stand-alone or in parallel to existing counter-measures.
FEDARS: Feature Extraction from Deep learning Architectures for face Recognition Systems
About this grant
Roles: Implementation
FEDARS studied feature extraction from deep learning architectures for face recognition, together with the industrial partner KeyLemon. It was our first attempt at building a face recognition system on deep learning rather than on hand-engineered features.
TABULA RASA: Trusted Biometrics under Spoofing Attacks
About this grant
Roles: Dataset acquisition (WP2), implementation (WP4)
Identity management for persons using biometrics has become a reality mainly because of the biometric passport but also because of the presence of more and more biometricenabled applications for personal computers. Although, the market for identity management using biometrics is dominated by some key players focusing mainly on high security applications, new exploitation routes are currently explored by SMEs. This market of identity management using biometrics is thus growing rapidly. Unfortunately, it has been shown recently that conventional biometric techniques, such as fingerprint or face recognition, are vulnerable to attacks. Generally, two types of attacks are considered: (1) indirect attacks due to intruders, such as cyber-criminal hackers, and (2) direct (spoofing) attacks, where a person tries masquerade as another one by falsifying data and thereby gaining an illegitimate advantage. Currently, spoofing attacks are a major problem for companies willing to market identity management solutions based on biometric technologies. Thus there is need for efficient and reliable solutions for detecting and circumventing these direct attacks. The typical countermeasure to a spoofing attack liveness detection that aims at detecting physiological signs of life. Another possible countermeasure is multi-modal biometrics. Indeed, voice recognition for instance could be performed jointly to face recognition and would be more robust to an attack on the video stream. Similarly, gait, face and iris recognition could be performed jointly. Additionally, it has been shown recently in pioneering work that emerging biometrics such as gait, vein or electro-physiological signals are potentially very difficult or impossible to spoof. The project should investigate as well such novel biometrics and further explore their advantages and limits. Hence, the goal of this proposal is to research, develop, evaluate and transfer antispoofing solutions.