Train yourself on the AV 2.0 Revolution
Autonomous Driving is evolving faster than you can catch up. One year, the revolution is Chat-GPT, the next one, everybody's already taken the idea into Foundation Models and End-To-End Learning.
If you'd like to stay up to date, there are 4 or 5 core ideas to understand:
- Transformers: and in particular Vision Transformers
- Self-Supervised Learning: How to train on "the entire internet" without supervision
- HydraNets: Multi-Task Learning techniques creating the new Foundation Models
- Neural Strike: Optimize & Deploy models to production
- Bird-Eye View/Occupancy: The core vision tasks
- End-To-End Learning: The core approach used in today's revolution, using a single network from sensors to output trajectory








