Professor Profile
Tenure-Track Researcher/Professor · Scientific Director of the AI4ILF Research Group ·
- Tenure-Track Researcher / Professor — Department of Mathematics and Computer Science (DMI), University of Catania, Computer Science [INFO-01/A].
- National Scientific Qualification (ASN) as Full Professor in both 01/B1 — Informatica and 09/H1 — Sistemi di Elaborazione delle Informazioni.
- Scientific Director of the AI4Industry, Legal and Financial (AI4ILF) research group @ IPLAB, University of Catania.
- Principal Investigator for the University of Catania of the European project NeAIxt (HORIZON-JU-Chips-2024-1-IA-T1, GAP-101194172); P.I. for STMicroelectronics of NEUROKIT2E and EdgeAI-Trust.
- Associate Editor of the IEEE Open Journal of the Computer Society (Q1); former Associate Editor of IET Networks and IET Image Processing.
- Formerly Senior Technical Staff Manager at the R&D Division of STMicroelectronics, Catania.
- World's Top 2% Scientists — sub-field Artificial Intelligence & Image Processing; author of 130+ scientific contributions and several international patents.
- Program Chair and co-organizer of workshops at CVPR, ICCV and ECCV — including PHAROS-AFE-AIMI @ ICCV 2025 and DEF-AI-MIA @ CVPR 2024.
- Member of the Ph.D. Scientific Boards in Computer Science (University of Catania) and in the National Ph.D. in Artificial Intelligence “Health and Life Science” (Università Campus Bio-Medico, Rome).
Scientific Visions
Small windows into my research ecosystem
A compact research area inspired by the five pillars of the AI4ILF group: artificial intelligence, hyperbolic spaces, medical AI, industrial intelligence, and financial AI systems.
Artificial Intelligence
Bio-inspired computational models, neuro-modulation, continual learning, adaptive Jacobian regularization, and knowledge distillation.
Hyperbolic Spaces
Deep learning in non-conventional geometries for hierarchical structure, Riemannian optimization, and stable curved embeddings.
Medical AI
Radiomics and CT-driven immunotherapy response prediction, MRI lesion segmentation, retinopathy assessment, and explainable clinical AI.
AI for Industry
Silicon-Carbide power device health monitoring, wafer defect map assessment, predictive maintenance, and edge inference for electric vehicles.
AI for Finance
Attractor-aware reinforcement learning for FX markets, Lyapunov–entropy dynamics, insolvency prediction, and legal document intelligence.
Research Agenda
Deep learning that is geometrically grounded, numerically stable, and deployable
The AI4ILF group develops theory, algorithms and industrial-grade software that unify geometric principles with modern deep learning — targeting inference on the spaces where the phenomena truly live, while preserving interpretability, robustness and edge efficiency.
Non-Euclidean Deep Learning
Hyperbolic and spherical manifolds, geodesic layers, Riemannian optimization, and hierarchical representation learning.
Bio-Inspired & Neuro-Modulated Models
Dopamine-inspired modulation, stability–plasticity balance, extended EWC, and Lipschitz-regularized continual learning.
Perceptive AI for Industry
Automotive and semiconductor applications: SiC power devices, domain adaptation, generative layers, and explainable screening.
Knowledge Distillation at the Edge
Compressing AI architectures onto resource-constrained microcontrollers and automotive-grade devices without losing reliability.
Selected Publications
Recent papers
The five most recent contributions. The complete list is available on Google Scholar and Scopus.
HypDeformNet: Edge-Deployable Deep Architecture with Jacobian-Stable Hyperbolic Deformation and Lipschitz Distillation for Immunotherapy Response Prediction
Rundo, F., Spata, M.O., Banna, G.L., Battiato, S. IEEE International Conference on Pattern Recognition (ICPR 2026), Springer LNCS vol. 16819.
DOI BibTeXStability–Plasticity Inspired Knowledge Distillation Expert System with Lipschitz-Regularized Neuro-Modulation for Silicon-Carbide Power Modules Health Monitoring in Next-Generation Electric Vehicles
Rundo, F., Spata, O.M., Pino, C., Calabretta, M., Messina, A., Rundo, M.S., Battiato, S. Elsevier Expert Systems with Applications, 132490.
DOI BibTeXResearch Projects
Funded programs and collaborative initiatives
HORIZON-JU-Chips-2024-1-IA-T1 · Principal Investigator for UniCT
NeAIxt — Next Generation of Edge AI Crossing Technology Fields
Grant Agreement GAP-101194172. Advancing edge-AI enablers and next-generation embedded Phase Change Memory (ePCM) on 18nm FD-SOI for secure, low-power European microcontrollers with in-memory computing.
HORIZON-KDT-JU-2023-1-IA · P.I. for STMicroelectronics
EdgeAI-Trust
Proposal 101139892-1. Trustworthy, decentralized edge intelligence across the European electronic components and systems value chain.
HORIZON-KDT-JU-2021-2-RIA · P.I. for STMicroelectronics
NEUROKIT2E
Proposal 101112268. Open-source deep learning toolkits and neuromorphic-oriented frameworks for European embedded AI.
HORIZON-KDT-JU · Research Team
R-PODID & ARCHIMEDES
Proposals 101097300 and 101112295. AI-driven reliability, predictive diagnostics and intelligent monitoring of power devices for electrified mobility.
H2020-ECSEL-2017-1-IA · Research Team
REACTION — First and European SiC Eight-Inches Pilot Line
Grant Agreement 783158. Deep learning for wafer monitoring, defect map assessment and lifetime estimation in Silicon-Carbide manufacturing.
PON R&I ARS01_00459 · 2014–2020
ADAS+ — Advanced Driver Assistance Systems
Perceptive deep learning for driver attention and drowsiness monitoring from PPG/ECG signals and low frame-rate video, without dedicated biometric sensors.
Further programs
ECS4DRES · HiCONNECTS · GAN4AP · ASTONISH · SATURN · Graphene Flagship Core 3
Additional European and national initiatives on heterogeneous integration, smart manufacturing, optical sensing for health, and advanced materials.
Teaching
Courses, supervision, and academic mentoring
Teaching at the Department of Mathematics and Computer Science, University of Catania, combined with doctoral training and invited lectures on perceptive deep learning, generative AI and medical imaging. Member of the Computer Science Ph.D. Scientific Board (UniCT) and of the National Artificial Intelligence Ph.D. Scientific Board — Università “Campus Bio-Medico” di Roma.
Medical Imaging
Deep learning for biomedical image analysis, radiomics, explainability and clinical decision support.
Reti di Calcolatori
Computer networks: architectures, protocol stacks, and network programming fundamentals.
Ulteriori Attività Formative
Supplementary training activities on applied AI and industrial research methodology.
Reti di Calcolatori — Channel M–Z
Computer networks for the Computer Science bachelor programme, channel M–Z.
Principi di Matematica, Informatica e Fisica
Foundations of mathematics, computer science and physics (channels 3 and 4).
Sistemi di Elaborazione delle Informazioni
Information processing systems for the degree course in Environmental and Natural Sciences. Editions: 2021/2022 · 2020/2021.
All teaching assignments from A.Y. 2022/2023 to date
Complete and always up-to-date list of courses on the official DMI faculty page (tab “Insegnamenti”).
Invited lectures and Ph.D. courses
University of Bologna, University of Padova, University of Calabria, AEIT Catania, and the II Level Master on molecular imaging and radiopharmaceuticals (UniCT).
Open topics for students
RAG embedding and dynamic indexing for LLM engines, intelligent FUOTA for EV ECUs, foundation models for agrifood monitoring, fully local LLM engines, hyperbolic manifolds in knowledge distillation on microcontrollers, and hierarchical neuro-modulation.
Curriculum Vitae
Academic appointments, honors, editorial service, and leadership
M.Sc. in Computer Science Engineering and Ph.D. in “Applied Mathematics for Technology”, University of Catania. Formerly Senior Technical Staff Manager in the R&D Division of STMicroelectronics, Catania. Program Chair and co-organizer of workshops at CVPR, ICCV and ECCV; member of IPLAB and of CVPL, the Italian association for research in Computer Vision, Pattern Recognition and Machine Learning.
Download Full CVContact
Prospective students, collaborators, and visiting scholars are welcome
Students interested in thesis or internship topics on deep learning for industrial, automotive, legal or financial applications — often in collaboration with companies — can write to arrange a face-to-face meeting in Room 361. Please include a brief research statement, CV, and links to publications or code.
francesco.rundo@unict.it · +39 095 7383046 · Cittadella Universitaria, Viale Andrea Doria 6, 95125 Catania, Italy