We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:


Current browse context:


Change to browse by:


References & Citations

DBLP - CS Bibliography


(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Robotics

Title: SMET: Scenario-based Metamorphic Testing for Autonomous Driving Models

Authors: Haiyang Ao, Ya Pan
Abstract: To improve the security and robustness of autonomous driving models, this paper presents SMET, a scenariobased metamorphic testing tool for autonomous driving models. The metamorphic relationship is divided into three dimensions (time, space, and event) and demonstrates its effectiveness through case studies in two types of autonomous driving models with different outputs.Experimental results show that this tool can well detect potential defects of the autonomous driving model, and complex scenes are more effective than simple scenes.
Comments: This article has been withdrawn by arXiv administrators due to excessive verbatim text overlap from external sources
Subjects: Robotics (cs.RO)
Cite as: arXiv:2104.01093 [cs.RO]
  (or arXiv:2104.01093v2 [cs.RO] for this version)

Submission history

From: arXiv Admin [view email]
[v1] Fri, 2 Apr 2021 15:39:40 GMT (403kb,D)
[v2] Thu, 8 Apr 2021 13:33:19 GMT (0kb,I)

Link back to: arXiv, form interface, contact.