Rethinking Self-driving: Multi-task Knowledge for Better Generalization and Accident Explanation Ability

By Toshiyuki Motoyoshi, Kazuma Sasaki, Tetsuya Ogata, Shigeki Sugano. 2018

42 citations recorded by the source corpus.

Current end-to-end deep learning driving models have two problems: (1) Poor generalization ability of unobserved driving environment when diversity of training driving dataset is limited (2) Lack of accident explanation ability when driving models don't work as expected. To tackle these two problems, rooted on the believe that knowledge of associated easy task is benificial for addressing difficult task, we proposed a new driving model which is composed of perception module for \textit{see and think} and driving module for \textit{behave}, and trained it with multi-task perception-related basic knowledge and driving knowledge stepwisely. Specifically segmentation map and depth map (pixel level understanding of images) were considered as \textit{what \& where} and \textit{how far} knowledge for tackling easier driving-related perception problems before generating final control commands for difficult driving task. The results of experiments demonstrated the effectiveness of multi-task perception knowledge for better generalization and accident explanation ability. With our method the average sucess rate of finishing most difficult navigation tasks in untrained city of CoRL test surpassed current benchmark method for 15 percent in trained weather and 20 percent in untrained weathers. Demonstration video link is: this https URL

Subfields: Artificial Intelligence, Automotive Engineering, Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Automotive Engineering

In Computer Vision and Pattern Recognition

Citations in the corpus

Multimodal End-to-End Autonomous Driving (2019, methodology): “Finally, MT stands for multitask learning, and the results are reproduced from [80].”
Trajectory-guided Control Prediction for End-to-end Autonomous Driving: A Simple yet Strong Baseline (2022, background): “Diverse optimized approaches are presented based on them, such as multi-modal inputs [22, 53], multi-task learning [56, 37, 24, 27, 31, 26, 63], dataset aggregation [45] and knowledge distillation [61, 60].”
End-to-End Autonomous Driving: Challenges and Frontiers (2023, background): “…as semantic segmentation [2], [11], [15], [28], [29], [31], [52], [139], [163], [208], [209], [210], [224], depth estimation [15], [28], [31], [208], [209], object detection [2], [28], [31], [52], affordance predictions [29], [208], [211], motion prediction [2], [52], and gaze map estimation [225].”
Exploring the Limitations of Behavior Cloning for Autonomous Driving (2019, methodology): “Nonetheless, recent works have proposed effective off-policy solutions, for instance by expanding the space of image/action pairs either using noise [22, 12], extra sensors [8], or modularization [37, 26, 5].”