Emperror: A Flexible Generative Perception Error Model for Probing Self-Driving Planners

To handle the complexities of real-world traffic, learning planners for self-driving from data is a promising direction. While recent approaches have shown great progress, they typically assume a setting in which the ground-truth world state is available as input. However, when deployed, planning ne...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:IEEE robotics and automation letters Jg. 10; H. 6; S. 5807 - 5814
Hauptverfasser: Hanselmann, Niklas, Doll, Simon, Cordts, Marius, Lensch, Hendrik P.A., Geiger, Andreas
Format: Journal Article
Sprache:Englisch
Veröffentlicht: Piscataway IEEE 01.06.2025
The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Schlagworte:
autonomous agents
Collision rates
Complexity theory
Decoding
Deep learning methods
Detectors
Failure modes
Heavily-tailed distribution
Noise
Noise measurement
object detection
Perception
Planning
segmentation and categorization
Three-dimensional displays
Trajectory
Transformers
ISSN:2377-3766, 2377-3766
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:To handle the complexities of real-world traffic, learning planners for self-driving from data is a promising direction. While recent approaches have shown great progress, they typically assume a setting in which the ground-truth world state is available as input. However, when deployed, planning needs to be robust to the long-tail of errors incurred by a noisy perception system, which is often neglected in evaluation. To address this, previous work has proposed drawing adversarial samples from a perception error model (PEM) mimicking the noise characteristics of a target object detector. However, these methods use simple PEMs that fail to accurately capture all failure modes of detection. In this letter, we present Emperror , a novel transformer-based generative PEM, apply it to stress-test an imitation learning (IL)-based planner and show that it imitates modern detectors more faithfully than previous work. Furthermore, it is able to produce realistic noisy inputs that increase the planner's collision rate by up to 85%, demonstrating its utility as a valuable tool for a more complete evaluation of self-driving planners.
Bibliographie:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:2377-3766
2377-3766
DOI:10.1109/LRA.2025.3562789