Autonomous Navigation through intersections with Graph ConvolutionalNetworks and Conditional Imitation Learning for Self-driving Cars
This addresses the problem of safe and efficient navigation in complex, dynamic environments for self-driving cars, representing an incremental improvement over existing methods.
The paper tackled autonomous navigation through unsignaled intersections by proposing a branched network combining graph convolutional networks and conditional imitation learning, resulting in higher success rates and shorter navigation times compared to baselines.
In autonomous driving, navigation through unsignaled intersections with many traffic participants moving around is a challenging task. To provide a solution to this problem, we propose a novel branched network G-CIL for the navigation policy learning. Specifically, we firstly represent such dynamic environments as graph-structured data and propose an effective strategy for edge definition to aggregate surrounding information for the ego-vehicle. Then graph convolutional neural networks are used as the perception module to capture global and geometric features from the environment. To generate safe and efficient navigation policy, we further incorporate it with conditional imitation learning algorithm, to learn driving behaviors directly from expert demonstrations. Our proposed network is capable of handling a varying number of surrounding vehicles and generating optimal control actions (e.g., steering angle and throttle) according to the given high-level commands (e.g., turn left towards the global goal). Evaluations on unsignaled intersections with various traffic densities demonstrate that our end-to-end trainable neural network outperforms the baselines with higher success rate and shorter navigation time.