Decoding Deep Learning: Implications for Low Power Systems and Security

On Wednesday, 26 March 2025, the Artificial Intelligence and Data Analytics Network (AIDAN) at Oxford Brookes University hosted an engaging and insightful seminar titled “Decoding Deep Learning: Implications for Low-Power Systems and Security”.

The session explored how deep learning is shaping the future of low-power systems and influencing security in today’s rapidly evolving technological landscape.

Distinguished Guest Speaker

The seminar featured a keynote talk by Stephen Roberts, Royal Academy of Engineering/Man Group Professor of Machine Learning at the University of Oxford.

Professor Roberts leads Oxford’s Machine Learning Research Group and co-directs the EPSRC Centre for Doctoral Training in Autonomous Intelligent Machines and Systems. His research applies Bayesian statistics and information theory across diverse domains including astronomy, biology, finance, and sensor networks. He is also co-founder of Mind Foundry, an AI company established in 2016, and is a Fellow of the Royal Academy of Engineering and Professorial Fellow at Somerville College.

Key Themes and Discussions

During the seminar, Professor Roberts provided valuable insights into:

• The challenges of deploying deep learning models on low-power and resource-constrained devices
• The balance between computational efficiency and model performance
• Security implications of AI systems in embedded and distributed environments
• Emerging research directions in Bayesian machine learning and intelligent systems

The talk encouraged lively discussion, with attendees engaging in thoughtful questions around sustainability, robustness, and the responsible deployment of AI technologies.