Transferring Autonomous Driving Knowledge on Simulated and Real Intersections
arXiv:1712.01106
Abstract
We view intersection handling on autonomous vehicles as a reinforcement learning problem, and study its behavior in a transfer learning setting. We show that a network trained on one type of intersection generally is not able to generalize to other intersections. However, a network that is pre-trained on one intersection and fine-tuned on another performs better on the new task compared to training in isolation. This network also retains knowledge of the prior task, even though some forgetting occurs. Finally, we show that the benefits of fine-tuning hold when transferring simulated intersection handling knowledge to a real autonomous vehicle.
Appeared in Lifelong Learning Workshop @ ICML 2017. arXiv admin note: text overlap with arXiv:1705.01197
References in corpus (8)
- Playing Atari with Deep Reinforcement Learning
- How transferable are features in deep neural networks?
- End to End Learning for Self-Driving Cars
- An Empirical Investigation of Catastrophic Forgetting in Gradient-Based Neural Networks
- CNN Features off-the-shelf: an Astounding Baseline for Recognition
- Safe, Multi-Agent, Reinforcement Learning for Autonomous Driving
- Domain Randomization for Transferring Deep Neural Networks from Simulation to the Real World
- A Deep Hierarchical Approach to Lifelong Learning in Minecraft
Cited by in corpus (4)
- Improving Generalization of Transfer Learning Across Domains Using Spatio-Temporal Features in Autonomous Driving
- Data-Driven Transferred Energy Management Strategy for Hybrid Electric Vehicles via Deep Reinforcement Learning
- Driving Tasks Transfer in Deep Reinforcement Learning for Decision-making of Autonomous Vehicles
- Strawberry Detection using Mixed Training on Simulated and Real Data