CausalAF: Causal Autoregressive Flow for Safety-Critical Driving Scenario Generation
This addresses the challenge of evaluating robustness in autonomous driving by improving scenario generation efficiency, though it appears incremental as it builds on existing flow-based methods with causal enhancements.
The paper tackles the problem of generating diverse and efficient safety-critical driving scenarios for evaluating autonomous driving systems by integrating causality as a prior into a flow-based generative framework, resulting in significantly fewer optimization resources needed for effective generation.
Generating safety-critical scenarios, which are crucial yet difficult to collect, provides an effective way to evaluate the robustness of autonomous driving systems. However, the diversity of scenarios and efficiency of generation methods are heavily restricted by the rareness and structure of safety-critical scenarios. Therefore, existing generative models that only estimate distributions from observational data are not satisfying to solve this problem. In this paper, we integrate causality as a prior into the scenario generation and propose a flow-based generative framework, Causal Autoregressive Flow (CausalAF). CausalAF encourages the generative model to uncover and follow the causal relationship among generated objects via novel causal masking operations instead of searching the sample only from observational data. By learning the cause-and-effect mechanism of how the generated scenario causes risk situations rather than just learning correlations from data, CausalAF significantly improves learning efficiency. Extensive experiments on three heterogeneous traffic scenarios illustrate that CausalAF requires much fewer optimization resources to effectively generate safety-critical scenarios. We also show that using generated scenarios as additional training samples empirically improves the robustness of autonomous driving algorithms.