Jia Cai

VLN-AVP achieves over 25% higher success rates in autonomous valet parking by leveraging zero-shot navigation and hybrid memory systems, redefining scalability in unseen environments.

History-guided control in DFP allows for adaptable motion planning that avoids the pitfalls of static pattern copying, leading to safer and more stable driving trajectories.

Forget trying to wrangle dynamic 4D scenes with recurrent networks – DynamicVGGT achieves state-of-the-art reconstruction accuracy using a surprisingly effective feed-forward approach.