Centralizing our Research Resources 
We are pleased to announce that the AVS Lab is now on Hugging Face. As our research increasingly focuses on integrating modern AI methods (including Vision-Language Models (VLMs) and Vision-Language-Action (VLA) architectures) into autonomous systems, we want to ensure our work remains easily accessible to the community. This new organization provides a unified, standardized platform to share our code, model weights, and datasets.

First Releases 
We released several datasets (accepted at ECCV 2026, ITSC 2026, etc.) alongside our first model checkpoints. Feel free to check out our research and contributions. If you have active questions, engage through the organization and the corresponding Hugging Face authors!

Future Updates Moving forward, we will continuously update the organization with new models and data from our ongoing projects in autonomous decision-making and motion planning. We invite researchers, students, and collaborators to visit the page and utilize the provided resources for their own work.

You can find and follow the AVS Lab Hugging Face organization here!