How would you build a scalable annotation workflow for large autonomous driving datasets for Autonomous Solutions?
