Learning End-to-end Autonomous Driving using Guided Auxiliary Supervision
This work addresses autonomous driving for urban settings, but it is incremental as it builds on existing imitation learning methods with auxiliary tasks.
The paper tackles the problem of learning autonomous driving in stochastic urban environments by proposing a Multi-task Learning from Demonstration (MT-LfD) framework that uses auxiliary supervision to guide driving command prediction, resulting in faster learning, better performance, and increased transparency.
Learning to drive faithfully in highly stochastic urban settings remains an open problem. To that end, we propose a Multi-task Learning from Demonstration (MT-LfD) framework which uses supervised auxiliary task prediction to guide the main task of predicting the driving commands. Our framework involves an end-to-end trainable network for imitating the expert demonstrator's driving commands. The network intermediately predicts visual affordances and action primitives through direct supervision which provide the aforementioned auxiliary supervised guidance. We demonstrate that such joint learning and supervised guidance facilitates hierarchical task decomposition, assisting the agent to learn faster, achieve better driving performance and increases transparency of the otherwise black-box end-to-end network. We run our experiments to validate the MT-LfD framework in CARLA, an open-source urban driving simulator. We introduce multiple non-player agents in CARLA and induce temporal noise in them for realistic stochasticity.