Computer Science > Robotics
[Submitted on 3 Mar 2021 (v1), revised 5 Mar 2021 (this version, v2), latest version 7 Jul 2021 (v4)]
Title:Worsening Perception: Real-time Degradation of Autonomous Vehicle Perception Performance for Simulation of Adverse Weather Conditions
View PDFAbstract:Autonomous vehicles rely heavily upon their perception subsystems to see the environment in which they operate. Unfortunately, the effect of varying weather conditions presents a significant challenge to object detection algorithms, and thus it is imperative to test the vehicle extensively in all conditions which it may experience. However, unpredictable weather can make real-world testing in adverse conditions an expensive and time consuming task requiring access to specialist facilities, and weatherproofing of sensitive electronics. Simulation provides an alternative to real world testing, with some studies develo** increasingly visually realistic representations of the real world on powerful compute hardware. Given that subsequent subsystems in the autonomous vehicle pipeline are unaware of the visual realism of the simulation, when develo** modules downstream of perception the appearance is of little consequence - rather it is how the perception system performs in the prevailing weather condition that is important. This study explores the potential of using a simple, lightweight image augmentation system in an autonomous racing vehicle - focusing not on visual accuracy, but rather the effect upon perception system performance. With minimal adjustment, the prototype system developed in this study can replicate the effects of both water droplets on the camera lens, and fading light conditions. The system introduces a latency of less than 8 ms using compute hardware that is well suited to being carried in the vehicle - rendering it ideally suited to real-time implementation that can be run during experiments in simulation, and augmented reality testing in the real world.
Submission history
From: Andrew Bradley [view email][v1] Wed, 3 Mar 2021 23:49:02 UTC (1,304 KB)
[v2] Fri, 5 Mar 2021 20:41:10 UTC (1,304 KB)
[v3] Mon, 31 May 2021 23:22:55 UTC (1,486 KB)
[v4] Wed, 7 Jul 2021 19:04:07 UTC (1,481 KB)
References & Citations
Bibliographic and Citation Tools
Bibliographic Explorer (What is the Explorer?)
Litmaps (What is Litmaps?)
scite Smart Citations (What are Smart Citations?)
Code, Data and Media Associated with this Article
CatalyzeX Code Finder for Papers (What is CatalyzeX?)
DagsHub (What is DagsHub?)
Gotit.pub (What is GotitPub?)
Papers with Code (What is Papers with Code?)
ScienceCast (What is ScienceCast?)
Demos
Recommenders and Search Tools
Influence Flower (What are Influence Flowers?)
Connected Papers (What is Connected Papers?)
CORE Recommender (What is CORE?)
arXivLabs: experimental projects with community collaborators
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.