Real-time Multi-target Path Prediction and Planning for Autonomous Driving aided by FCN
This addresses a key bottleneck in autonomous driving by enabling faster multi-target planning, though it is incremental as it combines existing methods (FCN and A*) rather than introducing a new paradigm.
The paper tackles the problem of real-time multi-target path planning for autonomous driving, which is limited by slow search-based methods in complex environments, and achieves real-time planning by using a fully convolutional neural network to predict path regions that constrain A* search, resulting in planning for 3 targets in less than 100ms.
Real-time multi-target path planning is a key issue in the field of autonomous driving. Although multiple paths can be generated in real-time with polynomial curves, the generated paths are not flexible enough to deal with complex road scenes such as S-shaped road and unstructured scenes such as parking lots. Search and sampling-based methods, such as A* and RRT and their derived methods, are flexible in generating paths for these complex road environments. However, the existing algorithms require significant time to plan to multiple targets, which greatly limits their application in autonomous driving. In this paper, a real-time path planning method for multi-targets is proposed. We train a fully convolutional neural network (FCN) to predict a path region for the target at first. By taking the predicted path region as soft constraints, the A* algorithm is then applied to search the exact path to the target. Experiments show that FCN can make multiple predictions in a very short time (50 times in 40ms), and the predicted path region effectively restrict the searching space for the following A* search. Therefore, the A* can search much faster so that the multi-target path planning can be achieved in real-time (3 targets in less than 100ms).