Embedding Synthetic Off-Policy Experience for Autonomous Driving via Zero-Shot Curricula

by   Eli Bronstein, et al.

ML-based motion planning is a promising approach to produce agents that exhibit complex behaviors, and automatically adapt to novel environments. In the context of autonomous driving, it is common to treat all available training data equally. However, this approach produces agents that do not perform robustly in safety-critical settings, an issue that cannot be addressed by simply adding more data to the training set - we show that an agent trained using only a 10 on the entire dataset. We present a method to predict the inherent difficulty of a driving situation given data collected from a fleet of autonomous vehicles deployed on public roads. We then demonstrate that this difficulty score can be used in a zero-shot transfer to generate curricula for an imitation-learning based planning agent. Compared to training on the entire unbiased training dataset, we show that prioritizing difficult driving scenarios both reduces collisions by 15 evaluation, all while using only 10


page 8

page 14

page 15


Hierarchical Model-Based Imitation Learning for Planning in Autonomous Driving

We demonstrate the first large-scale application of model-based generati...

Parallelized and Randomized Adversarial Imitation Learning for Safety-Critical Self-Driving Vehicles

Self-driving cars and autonomous driving research has been receiving con...

Brain-Inspired Deep Imitation Learning for Autonomous Driving Systems

Autonomous driving has attracted great attention from both academics and...

Randomized Adversarial Imitation Learning for Autonomous Driving

With the evolution of various advanced driver assistance system (ADAS) p...

Experience-Based Heuristic Search: Robust Motion Planning with Deep Q-Learning

Interaction-aware planning for autonomous driving requires an exploratio...

Zero-shot Transfer Learning of Driving Policy via Socially Adversarial Traffic Flow

Acquiring driving policies that can transfer to unseen environments is c...

Experience Filter: Using Past Experiences on Unseen Tasks or Environments

One of the bottlenecks of training autonomous vehicle (AV) agents is the...

Please sign up or login with your details

Forgot password? Click here to reset