GAD-Generative Learning for HD Map-Free Autonomous Driving
This work addresses the challenge of handling corner cases in autonomous driving for urban scenes, offering a potential commercial solution, though it appears incremental as it builds on existing deep-learning techniques.
The authors tackled the performance bottleneck in autonomous driving systems caused by rule-based planning and control by introducing a deep-learning-based approach that integrates prediction, decision, and planning modules. The method, trained on 10 hours of human driver data, supports all mass-production ADAS features and was successfully deployed on a test car without hardware modifications.
Deep-learning-based techniques have been widely adopted for autonomous driving software stacks for mass production in recent years, focusing primarily on perception modules, with some work extending this method to prediction modules. However, the downstream planning and control modules are still designed with hefty handcrafted rules, dominated by optimization-based methods such as quadratic programming or model predictive control. This results in a performance bottleneck for autonomous driving systems in that corner cases simply cannot be solved by enumerating hand-crafted rules. We present a deep-learning-based approach that brings prediction, decision, and planning modules together with the attempt to overcome the rule-based methods' deficiency in real-world applications of autonomous driving, especially for urban scenes. The DNN model we proposed is solely trained with 10 hours of human driver data, and it supports all mass-production ADAS features available on the market to date. This method is deployed onto a Jiyue test car with no modification to its factory-ready sensor set and compute platform. the feasibility, usability, and commercial potential are demonstrated in this article.