Jingbo Wang
Jingbo Wang is a research scientist at the Shanghai Artificial Intelligence Laboratory specializing in embodied AI and humanoid robotics. He completed his PhD at The Chinese University of Hong Kong in 2023 under Professor Dahua Lin, with research focused on physics-based simulation, embodied learning, and sim-to-real transfer for generalizable humanoid intelligence. He is best known for work including the Behavior Foundation Model (BFM) for humanoid robots and contributions to humanoid-object interaction research, and was recognized with the WAIC 2025 Rising Star Award.
“Imagine2Real: Towards Zero-shot Humanoid-Object Interaction via Video Generative Priors”
Source→“The primary bottleneck is the lack of high-fidelity 3D interaction data for humanoid robots, which prevents the learning of robust and versatile interaction policies.”
Source→“The Direct baseline achieves the lowest hands and base errors because it aggressively optimizes for point matching without any physical constraints. However, this comes at the cost of extreme jittering... making it nearly impossible to deploy on a real robot.”
Source→AI-extracted from podcast / newsletter / paper summaries. May contain errors.