The Designing Interaction Freedom via Active Inference (DIFAI) project is an ERC Advanced Grant, led by Rod Murray-Smith and which started on 1st January 2024. The project was selected by the ERC (proposal 101097708) and was funded by the UK Horizon guarantee scheme as EPSRC project EP/Y029178/1.

Problem: Reliable design of interactive systems using advanced sensors and machine learning (ML) is an unsolved problem. New sensors could expand how we interact with computers, but are still hard to design for, without overly constraining user behaviour. AI algorithms can reduce human workload, but can fail in complex contexts and can control, deskill and dis-empower people. We have no principled workflows for designing interaction to allow users to flexibly share autonomy with supporting AI.

Objectives: Integrate Active Inference theory into the human-computer interaction loop, linking human behaviour via sensors and ML/inference embeddings with dynamic mediating mechanisms to create end-to-end mutually adaptive loops between humans and systems. Develop novel interaction mechanisms for explicit and implicit control of AI autonomy levels to empower people via shared autonomy, while maintaining their agency. Create systematic, composable software tools for computational interaction design which support prototyping and analysis of ML-infused sensors coupled with humans, which can integrate probabilistic causal models to solve inverse problems with advanced sensors, adapting to closed-loop data.

Impact: Using ML to give users freedom to express themselves individually, we can be robust to user heterogeneity, ensure fairness for diverse users and enable creative uses of technologies. Our tools will form the foundation of future usable interfaces with novel sensors and rich data spaces, and advances can be shared and rapidly built on, transforming HCI research workflows. Applications: 1. Whole hand touch interaction via soft, programmable

News

• Review paper on Active inference in HCI released as arXiv preprint. https://arxiv.org/abs/2412.14741
• Two NeurIPS papers accepted and presented at NeurIPS2024, Is One GPU Enough? Pushing Image Generation at Higher-Resolutions with Foundation Models and Generative Fractional Diffusion Models.
• Our paper Towards Interaction Design with Active Inference: A Case Study on Noisy Ordinal Selection was accepted at the 5th International Workshop on Active Inference, 2024.
• Our paper SIM2VR: Towards Automated Biomechanical Testing in VR was accepted at UIST 2024.
• The University of Glasgow has generously opened the studentships' availablity to non-UK candidates. Aegean studentship application deadline has been extended to 30th August 2024.
• A second PhD studentship on optimal mechanism design (TBA).
• A new open PhD studentship with Aegean Airlines (30th October 2024 deadline)
• PDRA position appointed, which started in March 2024.
• 37 month post-doctoral research assistant post has gone live! Application deadline is 30th September 2023 (posted 1st August 2023).
• Project has been successfully transferred to the UK Horizon Guarantee Scheme from EPSRC (project EP/Y029178/1.) (20th June 2023).
• Project web page difai-project.org and twitter account @ERCdifai launched (21st May 2023).

Publications

• Preprints
  • R. Murray-Smith, J. H. Williamson, S. Stein, Active Inference and Human-Computer Interaction, https://arxiv.org/abs/2412.14741

• Papers in Journals
  • C. Kaul, K. J. Mitchell, K. Kassem, A. Tragakis, V. Kapitany, I. Starshynov, F. Villa, R. Murray-Smith, D. Faccio, AI-Enabled Sensor Fusion of Time of Flight Imaging and mmWave for Concealed Metal Detection, Sensors, 24 (18), 5865, 2024. pdf

• Papers in Conferences
  • S. Stein, J. H. Williamson, R. Murray-Smith, Towards Interaction Design with Active Inference: A Case Study on Noisy Ordinal Selection, International Workshop on Active Inference (IWAI), Oxford, 2024. pdf
  • A. Tragakis, M. Aversa, C. Kaul, R. Murray-Smith, Is One GPU Enough? Pushing Image Generation at Higher-Resolutions with Foundation Models, NeurIPS, 2024. pdf
  • F. Fischer, A. Ikkala, M. Klar, A. Fleig, M.Bachinski, R. Murray-Smith, P. Hämäläinen, A. Oulasvirta, J. Müller, SIM2VR: Towards Automated Biomechanical Testing in VR, UIST 2024. pdf

Vacancies

• Ph.D. studentship with Aegean Airlines. Model-based shared autonomy via active inference, Supervised by Sebastian Stein and Roderick Murray-Smith. Application deadline, 30th April 2025. Fully funded PhD studentship (fees + stipend at research council rates. For 2024-25 the stipend is £19237 per year. The stipend is usually non-taxable and does not need to be paid back.). This studentship is available to all candidates.
• Ph.D. studentship Optimal mechanism design and active inference for advanced human-computer interfaces to machine learning. Supervised by John Williamson and Roderick Murray-Smith. Application deadline, 30th July 2024. Fully funded PhD studentships (Fees + stipend at research council rates. For 2024-25 the stipend is £19237 per year. The stipend is usually non-taxable and does not need to be paid back.). This studentship is available to all candidates.

Team

Rod Murray-Smith is a professor in the School of Computing Science at the University of Glasgow, where he is a member of the Inference, Dynamics and Interaction group and works in the areas of human-computer interaction, machine learning, and control.	Rod Murray-Smith
John H. Williamson is a Senior Lecturer in Computing Science. His research interests are around machine learning for novel sensing devices, probabilistic modelling and filtering for interaction, and the use Bayesian methods more generally in HCI. In the past he has worked on brain-computer interfaces, mobile interaction and real-time sonification.	John H. Williamson
Sebastian Stein is a Research Fellow in the School of Computing Science at the University of Glasgow. His research interests are in intelligent interactive systems, spanning areas of HCI, ubiquitous computing, action recognition, computer vision and machine learning.	Sebastian Stein
Andrew Ramsay is a research associate in the Inference, Dynamics and Interaction group and his work typically fits the role of a “research software engineer”, assisting other researchers with a wide range of software development tasks in a variety of research projects.
Chaitanya Kaul is a post-doctoral research associate in the School of Computing Science, and applies Machine Learning to interesting problems in 3D Imaging, Computational Imaging, and Healthcare Applications, as well as understanding Privacy guarantees of Machine Learning models.
Markus Klar is a post-doctoral research associate in the School of Computing Science. His research interests are in simulating human movements during the interacction with computers, using dynamic biomechanical models, model predictive control, and deep reinforcement learning.	Markus Klar

Contact

Please contact Prof. Roderick Murray-Smith at roderick.murray-smith@glasgow.ac.uk