Recording published
Robotion · AI椰青
StarVLA：从 Vision-Language Model 到 Vision-Language-Action 的统一开源框架
A discussion of StarVLA’s modular design, connecting vision-language models with robot action learning and a unified research workflow.
Talk archive
Talks.
Written notes, slides, and videos from my talks and seminars.
2026
Recording published
ModelScope DevCon Nanjing · 具身智能分论坛 · Nanjing, China
StarVLA：构建开放，透明，灵活的具身开发链路
A presentation on an open, transparent, and flexible development workflow for embodied intelligence with StarVLA.
Let’s start a conversation.
For research talks and discussions on embodied intelligence and VLA models.