Invited Speakers

Prof. Tien-Tsin Wong
Monash University, Australia

Tien-Tsin Wong is Professor of Generative AI in the Department of Data Science and Artificial Intelligence at Monash University, Australia. Prior to this, he was a professor in the Computer Science & Engineering Department of the Chinese University of Hong Kong from 1999 to 2024. He received his B.Sc., M.Phil., and Ph.D. degrees in Computer Science from the Chinese University of Hong Kong in 1992, 1994, and 1998, respectively. He received the IEEE Transactions on Multimedia Prize Paper Award 2005 and the Young Researcher Award 2004. He has served an associate editors for IEEE Transactions on Visualization and Computer Graphics, Computer Graphics Forum, Computational Visual Media, and The Visual Computer.
He is known with his works in Generative AI for Video Generation, Computational Manga, Image-based Relighting, Ambient Occlusion (Dust Accumulation Simulation), Sphere Maps, and GPU techniques. His recent works include Dynamicrafter, Tooncrafter, MotionCanvas and VLIPP. More information about him can be found at https://ttwong12.github.io/

Speech Title: Towards Generative Filming: Progress and Challenges 

Abstract: The rapid advancement of generative AI has made the once-distant dream of creating entire films using artificial intelligence increasingly tangible. Traditional film production is costly, time-consuming, and dependent on extensive technical expertise in modeling, animation, and rendering. In contrast, generative models, particularly diffusion-based techniques, promise high-quality visual richness with drastically lower modeling effort and a reduced skill barrier, offering new creative possibilities for both professionals and amateurs.
This talk explores the emerging generative filming: using AI to generate, control, and refine visual narratives. I will review our team’s research work in controlling diffusion models for video generation, including work on image, and keyframe-based conditioning, as well as motion and camera trajectory control. We will also discuss fundamental challenges that must be overcome before AI-generated films can rival traditional production pipelines, such as physics plausibility, long-term content consistency, and the gap between 2D training data and our inherently 3D physical world. I will also discuss about a hybrid approach, combining the physical rigor of computer graphics with the generative power of AI, may be the most practical path forward toward fully AI-assisted filmmaking, at the current moment.