Present
CEO & Founder
Lead Sapiente Education’s strategy and research programmes, connecting students with expert mentors for personalised learning and original research projects.
Research, education & human-centred innovation
I am an AI researcher and entrepreneur based in Dubai. My work spans efficient deep learning, few-shot learning, Bayesian methods, computer vision, and robotics.
I am the founder and CEO of Sapiente Education, which connects students with expert mentors for personalised learning and research, and Vice-President of the Human Economic Forum, promoting dialogue on technology, ethics, and a human-centred economy.
Previously, I worked in AI research at Tools for Humanity and held postdoctoral research positions at the University of Cambridge and the University of Edinburgh.
Current positions
Present
Vice-President
Support the Human Economic Forum’s mission to build a human-centred economy, promoting dialogue on technology, ethics, and their impact on society.
Previous research & industry experience
2024–2025
Senior Artificial Intelligence Research Engineer
Tools for Humanity / Worldcoin
Developed deep neural networks for iris and face recognition, supporting biometric identity verification within the Artificial Intelligence and Biometrics team.
2021–2024
Senior Postdoctoral Researcher
Worked with Richard Turner on efficient machine learning through the EPSRC “Machine Learning for Tomorrow” project, in collaboration with Microsoft Research.
2018–2021
Postdoctoral Researcher
Worked with Amos Storkey on Bayesian meta-learning, few-shot learning, and deep kernels for Gaussian processes through a research collaboration with Huawei.
2018
Research Intern
Researched disentangled visual representations in deep autoencoders with Patrick Fox-Roberts and Edward Rosten in the Camera Platform team, London.
2012–2015
Robotics Engineer
Eurolink Systems
Developed autonomous control systems for ground robots and aerial vehicles, supporting applications in search and rescue, patrolling, and bomb disposal.
2011–2012
Research Intern
Laboratory of Artificial Life and Robotics (LARAL)
Studied neural networks and genetic algorithms for evolutionary robotics with Domenico Parisi, exploring how learning and evolution shape robotic behaviour.
Education
2015–2018
PhD in Machine Learning and Robotics
University of Plymouth
Research on social skills in humanoid robots using machine learning. Supervisors: Angelo Cangelosi, Torbjorn Dahl, and Giorgio Metta.
2009–2011
MSc in Neuroscience
Sapienza University of Rome
Supervisors: Stefano Puglisi Allegra, Domenico Parisi, and Gianluca Baldassarre.
2006–2009
BSc in Experimental Cognitive Psychology
Sapienza University of Rome
Supervisors: Marta Olivetti Belardinelli and Valerio Santangelo.
Selected publications
For the full and latest list, visit Google Scholar.
Torre, I., Zanatto, D., Patacchiola, M., Cangelosi, A., & Goslin, J. (2026). "The Right Voice for the Job: Voice and Gaze Cues in a Repeated Investment Game with a Robot". IEEE International Conference on Robot and Human Interactive Communication (RO-MAN), pp. 1009–1016. [DOI]
Bachèlery, M.-L., Brajard, J., Patacchiola, M., Illig, S., & Keenlyside, N. (2025). "Predicting Atlantic and Benguela Niño events with deep learning". Science Advances, 11(14), eads5185. [DOI]
Ochal, M., Patacchiola, M., Boudiaf, M., & Wang, S. (2024). "EOL: Transductive Few-Shot Open-Set Recognition by Enhancing Outlier Logits". arXiv preprint, arXiv:2408.02052. [arXiv]
Patacchiola, M., Shysheya, A., Hofmann, K., & Turner, R.E. (2024). "Transformer Neural Autoregressive Flows". Structured Probabilistic Inference & Generative Modeling (SPIG) Workshop at ICML. [arXiv] [GitHub]
Ochal, M., Patacchiola, M., Vazquez, J., Storkey, A., & Wang, S. (2023). "Few-Shot Learning With Class Imbalance". IEEE Transactions on Artificial Intelligence, 4(5), 1348–1358. [IEEE Xplore] [GitHub]
Patacchiola, M., Sun, M., Hofmann, K., & Turner, R.E. (2023). "Comparing the Efficacy of Fine-Tuning and Meta-Learning for Few-Shot Policy Imitation". Conference on Lifelong Learning Agents (CoLLAs), PMLR 232:878–908. [PMLR] [PDF]
Shysheya, A., Bronskill, J., Patacchiola, M., Nowozin, S., Turner, R.E. (2023). "FiT: Parameter Efficient Few-shot Transfer Learning for Personalized and Federated Image Classification". International Conference on Learning Representations (ICLR). [arXiv] [GitHub]
Patacchiola, M., Bronskill, J., Shysheya, A., Hofmann, K., Nowozin, S., Turner, R.E. (2022). "Contextual Squeeze-and-Excitation for Efficient Few-Shot Image Classification". Advances in Neural Information Processing Systems (NeurIPS). [arXiv] [GitHub]
Bronskill*, J., Massiceti*, D., Patacchiola*, M., Hofmann, K., Nowozin, S., & Turner, R.E. (2021). "Memory Efficient Meta-Learning with Large Images". Advances in Neural Information Processing Systems (NeurIPS). *Co-first authors. [arXiv]
Sendera, M., Tabor, J., Nowak, A., Bedychaj, A., Patacchiola, M., Trzcinski, T., Spurek, P., & Zieba, M. (2021). "Non-Gaussian Gaussian Processes for Few-Shot Regression". Advances in Neural Information Processing Systems (NeurIPS). [arXiv]
Ochal, M., Patacchiola, M., Storkey, A., Vazquez J., & Wang S. (2021). "How Sensitive are Meta-Learners to Dataset Imbalance?". Learning to Learn Workshop - International Conference on Learning Representations (ICLR) [arXiv] [GitHub]
Patacchiola, M., & Storkey, A. (2020). "Self-Supervised Relational Reasoning for Representation Learning". Advances in Neural Information Processing Systems (NeurIPS). Spotlight (top 3%) [arXiv] [GitHub] [Poster]
Patacchiola, M., Turner, J., Crowley, E. J., O'Boyle, M., & Storkey, A. (2020). "Bayesian Meta-Learning for the Few-Shot Setting via Deep Kernels". Advances in Neural Information Processing Systems (NeurIPS). Spotlight (top 3%) [arXiv] [GitHub] [Poster]
Antoniou, A., Patacchiola, M., Ochal, M., & Storkey, A. (2020). "Defining Benchmarks for Continual Few-Shot Learning". MetaLearn Workshop - Advances in Neural Information Processing Systems (NeurIPS) [arXiv] [GitHub] [YouTube] [Dataset]
Polvara*, R., Patacchiola*, M., Hanheide, M., & Neumann, G. (2020). "Sim-to-Real Quadrotor Landing via Sequential Deep Q-Networks and Domain Randomization". Robotics, 9(1), 8. *Co-first authors. [PDF] [GitHub]
Patacchiola, M., Fox-Roberts, P., Rosten, E. (2020). "Y-Autoencoders: Disentangling Latent Representations via Sequential Encoding". Pattern Recognition Letters, Elsevier, vol. 140, pp. 59-65. [arXiv] [DOI] [GitHub]
Thabet, M., Patacchiola, M., & Cangelosi, A. (2019). "Sample-efficient Deep Reinforcement Learning with Imaginary Rollouts for Human-Robot Interaction". IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) [arXiv]
Polvara* R., Patacchiola*, M., Sharma S., Wan J., Manning A., Sutton R., Cangelosi A. (2018). "Toward End-To-End Control for UAV Autonomous Landing Via Deep Reinforcement Learning". The 2018 International Conference on Unmanned Aircraft Systems (ICUAS). *Co-first authors. [PDF]
Surace, L., Patacchiola, M., Battini Sonmez, E., Spataro, W., & Cangelosi, A. (2017). "Emotion Recognition in the Wild using Deep Neural Networks and Bayesian Classifiers". In Proceedings of the Fifth Emotion Recognition in the Wild (EmotiW) Challenge, Glasgow, United Kingdom. [arXiv]
Patacchiola, M., & Cangelosi, A. (2017). "Head Pose Estimation in the Wild using Convolutional Neural Networks and Adaptive Gradient Methods". Pattern Recognition, Elsevier, vol. 71, pp. 132-143. [PDF] [DOI]
Patacchiola, M., & Cangelosi, A. (2016). "A Developmental Bayesian Model of Trust in Artificial Cognitive Systems". In Proceedings of the International Conference on Development and Learning and Epigenetic Robotics (ICDL-EpiRob). [PDF]
Zanatto, D., Patacchiola, M., Goslin, J., & Cangelosi, A. (2016). "Priming Anthropomorphism: Can the credibility of humanlike robots be transferred to non-humanlike robots?". In The Eleventh ACM/IEEE International Conference on Human-Robot Interaction (HRI). [PDF]