NuRisk: Visual Question Answering for Autonomous Driving Risk Assessment

NuRisk is a visual question answering dataset focusing on risk assessment for autonomous driving. Each row contains a BEV image, a driving-related question, and a ground truth answer. The dataset was created by Yuan-avs and was last updated on Hugging Face in May 2026.

Use Cases

• Training VQA models for scene understanding based on BEV imagery.
• Evaluating agent-level risk assessment algorithms for autonomous driving.
• Benchmarking multimodal reasoning systems on driving-specific questions and answers.

Strengths

• Focuses on the specific domain of risk assessment for autonomous driving.
• Provides a multimodal structure with images, questions, and answers.

Limitations

• Dataset size, column details, and sample data are unknown, limiting suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.

Provenance

Source

Yuan-avs on Hugging Face.

Freshness

Last updated 2026-05-29 12:33:22.