30 May 2023 — Held in conjunction with NUS Computing’s 25^th Anniversary, the inaugural Singapore Vision Day 2023 welcomed for the first time in Singapore over 170 computer vision researchers in academia and the industry to foster an exchange of research ideas and build a community of experts in computer vision. It was also an opportunity for computing students to get exposure to industry participants and to engage companies that plan to incorporate computer vision in their future R&D plans.

Singapore: AI hub for Southeast Asia
In his welcome address, Professor Tan Kian Lee, Dean of NUS Computing said, “Computer vision is a rapidly evolving field, dynamic and transformative, and has gained a lot of interest from around the world. It is revolutionizing many industries and has been seamlessly integrated into our daily lives. We see it being used in robots, autonomous vehicles, medical imaging, and much more. By bringing computer vision researchers across Singapore together, we hope to foster greater collaboration and ultimately aim to cultivate Singapore’s reputation as an Artificial Intelligence hub for Southeast Asia.”

Global speakers on Computer Vision
The event featured a series of keynote talks by overseas speakers including Jiajun Wu from Stanford University who spoke on “Understanding the Visual World Through Naturally Supervised Code” and Vincent Sitzmann from Massachusetts Institute of Technology (MIT) who presented on “3D Representation Learning at Scale”.

Or Litany a senior research scientist at Nvidia discussed an innovative neural simulation approach that learns to simulate driving scenarios from data. His key research focus covers three key areas: scene reconstruction in both appearance and geometry, motion generation of humans and vehicles, and LiDAR synthesis. Niloy Mitra from University College London (UCL) described results using synthetic and real data and how we can extend these models to produce high-quality photorealistic outputs. Finally, Simon Lucey from the University of Adelaide explored a new direction based on the concept of a “neural prior” that relies on no training dataset whatsoever. A neural prior speaks to the remarkable ability of neural networks to both memorise training and generalise unseen testing examples.

From Art to Autonomous Vehicles
To showcase how AI is adopted in the world of art, guest speaker Wyn-Lyn Tan, a visual artist described how she creates generative AI videos based on machine learning using her own physical paintings. Another guest speaker, Jia Cai, a technical expert in Autonomous Driving from Huawei Central Media Technology Institute spoke about his research in Visual perception of general obstacles, End-to-End Autonomous Driving and 3D reconstruction (high geometrical accuracy).

Lightning Talks by Singapore Academia
In addition, AI academia from other Singapore universities presented lightning talks on topics ranging from “Responsible Generative AI in Computer Vision” by Ngai Mun Cheung of SUTD to “Face Processing and Human Motion Analysis” by Tat-Jen Cham (NTU). Other speakers included Robby Tan (NUS);  Mengmi Zhang (NTU/ASTAR); Shengfeng HE (SMU); Guosheng Lin (NTU); Gim Hee Lee (NUS); Angela Yao (NUS); Malika Meghjani (SUTD); Chong Hwa Ngo (SMU); Mohan Kankanhalli (NUS); Na Zhao (SUTD); Ziwei Liu (NTU); Xinchao Wang (NUS); Peng Song (SUTD); Jun Liu (SUTD); Daniel Lin (SMU); Mike Shou (NUS). About 36 student researchers also presented posters of their research projects at the event.

The event was organized by Associate Professor Gim Hee Lee and Dean’s Chair Assistant Professor Angela Yao from NUS Computing and co-sponsored by NUS AiL, Huawei and NUS Computing.