Safer End-to-End Autonomous Driving via Conditional Imitation Learning and Command Augmentation

TL;DR

This paper introduces a novel probabilistic model for imitation learning in autonomous driving that enables command-conditional control without complex architectures, improving robustness and interpretability.

Contribution

It presents a disentangled probabilistic latent variable model for command-conditional steering, avoiding branched architectures and enhancing controller stability and safety.

Findings

Achieves command-conditional steering without branched networks.

Demonstrates improved stability and robustness of the driving controller.

Extends the model to handle unsafe commands, such as turning into walls.

Abstract

Imitation learning is a promising approach to end-to-end training of autonomous vehicle controllers. Typically the driving process with such approaches is entirely automatic and black-box, although in practice it is desirable to control the vehicle through high-level commands, such as telling it which way to go at an intersection. In existing work this has been accomplished by the application of a branched neural architecture, since directly providing the command as an additional input to the controller often results in the command being ignored. In this work we overcome this limitation by learning a disentangled probabilistic latent variable model that generates the steering commands. We achieve faithful command-conditional generation without using a branched architecture and demonstrate improved stability of the controller, applying only a variational objective without any domain-specific adjustments. On top of that, we extend our model with an additional latent variable and augment the dataset to train a controller that is robust to unsafe commands, such as asking it to turn into a wall. The main contribution of this work is a recipe for building controllable imitation driving agents that improves upon multiple aspects of the current state of the art relating to robustness and interpretability.