Multi-Modal Traffic Scenario Generation for Autonomous Driving System Testing
Autonomous driving systems (ADS) require extensive testing and validation before deployment. However, it is tedious and time-consuming to construct traffic scenarios for ADS testing. In this paper, we propose TrafficComposer, a multi-modal traffic scenario construction approach for ADS testing. TrafficComposer takes as input a natural language description of a desired traffic scenario and a complementary traffic scene image. Then, it generates the corresponding traffic scenario in a simulator, such as CARLA and LGSVL. Specifically, TrafficComposer integrates high-level dynamic information about the traffic scenario from the NL description and intricate details about the surrounding vehicles, pedestrians, and the road network from the image. The information from the two modalities is complementary to each other and helps generate high-quality traffic scenarios for ADS testing. On a benchmark of 120 traffic scenarios, TrafficComposer achieves 97.0% accuracy, outperforming the best-performing baseline by 7.3%. Both direct testing and fuzz testing experiments on six ADSs prove the bug detection capabilities of the traffic scenarios generated by TrafficComposer. These scenarios can directly discover 37 bugs and help two fuzzing methods find 33%–124% more bugs serving as initial seeds.
Mon 23 JunDisplayed time zone: Amsterdam, Berlin, Bern, Rome, Stockholm, Vienna change
10:30 - 12:30 | Autonomous DrivingDemonstrations / Research Papers / Industry Papers at Cosmos 3A Chair(s): Nassim Belmecheri | ||
10:30 20mTalk | On-Demand Scenario Generation for Testing Automated Driving Systems Research Papers Songyang Yan Xi'an Jiaotong University, Xiaodong Zhang Xidian University, Kunkun Hao Synkrotron, Inc., Haojie Xin Xi'an Jiaotong University, Yonggang Luo Chongqing Changan Automobile Co. Ltd, Jucheng Yang Chongqing Changan Automobile Co. Ltd, Ming Fan Xi'an Jiaotong University, Chao Yang Xidian University, Jun Sun Singapore Management University, Zijiang Yang University of Science and Technology of China and Synkrotron, Inc. DOI Pre-print | ||
10:50 20mTalk | Multi-Modal Traffic Scenario Generation for Autonomous Driving System Testing Research Papers Zhi Tu Purdue University, Liangkun Niu Purdue University, Wei Fan Purdue University, Tianyi Zhang Purdue University DOI Pre-print | ||
11:10 10mTalk | CCTest: Critical Configuration Testing for Autonomous Driving Systems Demonstrations Changwen Li , Joseph Sifakis University Grenoble Alpes; CNRS; Grenoble INP; VERIMAG, Rongjie Yan Institute of Software at Chinese Academy of Sciences; University of Chinese Academy of Sciences, Jian Zhang Institute of Software at Chinese Academy of Sciences; University of Chinese Academy of Sciences | ||
11:20 10mTalk | HazSim: An Urban Driving Simulator for Testing Perception Systems of ADSes Against Hazards Demonstrations Xiaodong Zhang Xidian University, Jie Bao Xi'an Jiaotong University, Yulong Shen Xidian University, Qin Xia Xi'an Jiaotong University, Zijiang Yang University of Science and Technology of China and Synkrotron, Inc. | ||
11:40 20mTalk | A Comprehensive Study of Bug-Fix Patterns in Autonomous Driving Systems Research Papers Yuntianyi Chen University of California, Irvine, Yuqi Huai University of California, Irvine, Yirui He University of California, Irvine, Shilong Li University of California, Irvine, Changnam Hong University of California, Irvine, Alfred Chen University of California, Irvine, Joshua Garcia University of California, Irvine DOI Pre-print | ||
12:00 10mTalk | PCLA: A Framework for Testing Autonomous Agents in the CARLA Simulator Demonstrations Masoud Jamshidiyan Tehrani Università della Svizzera italiana, Jinhan Kim Università della Svizzera italiana (USI), Paolo Tonella USI Lugano | ||
12:10 20mTalk | AutoTracer: A Low-Overhead Tracing Framework for Autonomous Driving System Industry Papers Bo Jiang Beihang University, Fancheng Shu SKLCCSE，Beihang University, Yuyang Cui SKLCCSE，Beihang University, Xiangjie Wang SKLCCSE，Beihang University, Peng Tang SKLCCSE，Beihang University, Lei Wang SKLCCSE，Beihang University, Weiping Zhang DiDi Global, Yong Wang Beihang University, Guobin Wu DiDi Global | ||
Cosmos 3A is the first room in the Cosmos 3 wing.
When facing the main Cosmos Hall, access to the Cosmos 3 wing is on the left, close to the stairs. The area is accessed through a large door with the number “3”, which will stay open during the event.