AI (Artificial Intelligence) and the
metaverse are today’s hottest topics. Jointly organised by HKBU’s Department of Computer Science and NVIDIA, the Symposium on AI, Metaverse and Beyond, held on Shaw Campus on December 2, 2022, attracted many academics and industry professionals who are enthusiastic about this irreversible trend.
The Symposium started with a public keynote on “Digital Twins: Clash of Mathematics, AI, Data and High Performance Simulation” by Professor Simon See, Global Head and Chief Solution Architect, NVIDIA AI Technology Center with simultaneous live streaming.
During the keynote talk, Professor See, also a Professor at Shanghai Jiao Tong University, Beijing University of Posts and Telecommunications and Universitas Indonesia, among others, discussed how mathematics, data, simulation and AI together make digital twins possible for different applications.
With
digital twins technology, and the right data and processes, Professor See said, scientists, engineers, biologists and economists can now gain better control and understanding over their resources and assets, and quickly identify the actions and strategies needed to achieve sustainable performance improvement. To further illustrate the power of digital twins, See highlighted how NVIDIA Omniverse will help scientists to predict climate change through NVIDIA
Earth-2, an initiative to build a digital twin of the Earth to accelerate climate research.
Professor See’s inspirational talk was followed by research presentations delivered by three scholars of the Department of Computer Science, HKBU, as well as several postgraduate students who joined the NVIDIA AI Technology Centre as student interns.
Dr. Jie Chen introduced his project on “Nature-Inspired Motion Generation and AI Choreography”, illustrating how natural and up-to-beat motion can be synthesised and how nature and physical laws can inspire choreography.
Dr. Renjie Wan presented a talk on “Building Museums in A Day” and shared his insights on how neural rendering techniques can ease the task of migrating exhibitions from offline to online using a collection of 2D images. He introduced the methods to alleviate the reflection issue of the captured images to ensure high-quality neural rendering.
Dr. Bo Han’s topic is on “Towards Trustworthy Learning and Reasoning under Noisy Data” and he introduced his work on trustworthy learning and reasoning from three human-inspired views, including reliability, robustness, and interaction.
The Symposium was also featured with presentations by research students where they received comments from Professor See as well as other experts in the field.
