"Dataset for Retrieval-Augmented OpenSCENARIO2 (OSC2) Scenario Generation and Comparative Evaluation in CARLA\u2013Autoware Simulation"
Abstract
"This dataset supports the manuscript entitled \u201cRetrieval-Augmented OpenSCENARIO2 (OSC2) Scenario Generation for Comparative Autonomous Vehicle Stress Testing\u201d submitted to IEEE Transactions on Intelligent Transportation Systems. It contains the complete Retrieval-Augmented Generation (RAG) database, generated OpenSCENARIO2 (OSC2) scenario artifacts, and structured evaluation results used in a CARLA\u2013Autoware co-simulation framework. The dataset includes reusable geometry, spawn, and behavior snippets that form the indexed RAG database, example scenarios generated via template-based, from-scratch LLM, and RAG-based approaches, as well as Safety Pool\u2013compatible conversions and aggregated SafeBench evaluation metrics. Additionally, it provides ablation study outputs, coverage analyses, and difficulty-controlled experiment logs that enable reproducibility of hallucination reduction, validity assessment, and diversity comparisons across generation strategies. The dataset is intended for simulation-based comparative evaluation and stress testing research in autonomous driving and does not constitute real-world safety validation."
Read the paper
Related studies
- Knowledge Management Perspective of Generative Artificial Intelligence
- Navigating Generative Artificial Intelligence Promises and Perils for Knowledge and Creative Work
- Characterizing the evolving research on enterprise content management
- Large language models and multimodal foundation models for precision oncology
- From Generative AI to Extended Reality: Multidisciplinary Perspectives on the Challenges, Opportunities, and Future of Educational Computing
- Generative Artificial Intelligence: A New Engine for Advancing Environmental Science and Engineering
Go deeper in Evidence Hub
Evidence Hub adds a study data sheet (design, sample size, evidence level, key numbers), an AI summary and a chat grounded in the paper's text.