DERI Seminar with Prof Yang Hao
When: Thursday, May 4, 2023, 11:00 AM - 12:00 PM
Where: Zoom
Speaker: Prof Yang Hao, QinetiQ/Royal Academy of Engineering Research Chair at QMUL
Zoom link: https://qmul-ac-uk.zoom.us/j/81148100921
Title: Machine Learning for Future Electromagnetics: Opportunities and Challenges
Abstract: The talk will provide an overview of ongoing research on how to apply machine learning to solve challenging problems related to electromagnetics. Three use cases will be presented:
I will start with a design of disordered surfaces with hundreds of thousands to overcome longstanding issues related to bandwidth and unwanted diffraction. Herein, an interactive learning approach is developed to build new meta-atom datasets which include the effect of mutual coupling. A deep learning-based model is developed to extract features of incident/reflection waves and their neighbouring interaction responses from a limited number of known meta-atoms. Finally, the deep neural network is incorporated with optimization algorithms to design, as an example, large-scale metasurfaces for beam manipulation and wideband scattering reduction.
The success of ML in materials property prediction depends heavily on how the materials are represented for learning. Here, a new concept of formula graph which unifies stoichiometry-only and structure-based material descriptors is introduced. A self-attention integrated GNN that assimilates a formula graph is further developed and it is found that the proposed architecture produces material embeddings transferable between the two domains. The model is applied in a challenging exemplar to predict the complex dielectric function of materials and nominate new substances that potentially exhibit epsilon-near-zero phenomena.
Finally, we apply ML to emotion states recognition using wireless signals. Currently, standoff emotion detection is mostly reliant on the analysis of facial expressions and/or eye movements acquired from optical or video cameras. In this talk, I will report a novel deep neural network (DNN) architecture based on the fusion of raw RF data and the processed RF signal for classifying and visualising various emotion states. Our results indicate that using wireless signals for stand-by emotion state detection is a better alternative to other technologies with high accuracy and have much wider applications in future studies of behavioural sciences.
Bio: Professor Yang Hao is QinetiQ/Royal Academy of Engineering Research Chair at Queen Mary University of London. He also serves in the management team of Cambridge Graphene Centre since 2013. Prof. Hao was the Editor-in-Chief for the IEEE ANTENNAS AND WIRELESS PROPAGATION LETTERS. He founded a new open access journal and is now the Editor-in-Chief of EPJ Applied Metamaterials.
His work has been recognised both nationally and internationally through his books “Antennas and Radio Propagation for Body-Centric Wireless Communications” and “FDTD Modeling of Metamaterials: Theory and Applications,” (Artech House, USA) and highly cited papers published in leading journals, including Nature Communications, Advanced Science, Physical Review Letters, Applied Physics Letters, IEEE Proceedings, and Transactions.
His research on transformation optics and metamaterials have led to many tangible benefits for a range of industrial products. One example is metalens antenna designs for satellite communications. This technology has been fully scoped and is currently commercialized under a startup of Isotropic System Limited (All.Space).
Prof. Hao won many accolades, including the prestigious AF Harvey Prize in 2015, the BAE Chairman’s Silver Award in 2014, and the Royal Society Wolfson Research Merit Award in 2013. He was the AdCom Member and currently serves as the Chair of Publication Committee for IEEE Antennas and Propagation Society. Prof. Hao is an elected Fellow of Royal Academy of Engineering, IEEE and IET.