English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Tools annual

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Autonomous driving (AD) systems are often built and tested in a modularfashion, where the performance of different modules is measured usingtask-specific metrics. These metrics should be chosen so as to capture thedownstream impact of each module and the performance of the system as a whole.For example, high perception quality should enable prediction and planning tobe performed safely. Even though this is true in general, we show here that itis possible to construct planner inputs that score very highly on variousperception quality metrics but still lead to planning failures. In an analogyto adversarial attacks on image classifiers, we call such inputs\textbf{adversarial perception errors} and show they can be systematicallyconstructed using a simple boundary-attack algorithm. We demonstrate theeffectiveness of this algorithm by finding attacks for two different black-boxplanners in several urban and highway driving scenarios using the CARLAsimulator. Finally, we analyse the properties of these attacks and show thatthey are isolated in the input space of the planner, and discuss theirimplications for AD system deployment and testing.

Attacking Motion Planners Using Adversarial Perception Errors