Recent Talks
Selected invited keynotes on autonomous driving, augmented reality, generative AI and responsible AI.
Simulations and generative AI for training and verification of autonomous driving systems.
Factorized inverse path tracing for material and emitter estimation in augmented reality.
VLMs for perception and planning, diffusion models for simulation and LLM agents for self-driving devops.
LLM agents and vision-language models for full-stack autonomous driving.
Shape recovery from single or few images for challenging specular or transparent objects.
Neural rendering and graphics-based vision for augmenting indoor architecture.
Physically-based understanding of videos for autonomous driving and augmented reality.
High-quality inverse rendering for shape, material and lighting in images.
Learning generalizable, interpretable and resposible representations for video applications.