scenario-generation

[Survey Paper] This repository collects research papers of large Foundation Models for Scenario Generation and Analysis in Autonomous Driving. The repository will be continuously updated to track the latest update.

Tools for Stochastic Simulation using diffusion models (R).

ORNL’s Real-Twin project is a streamlined scenario generation tool that automatically integrates real-world traffic data to create high-fidelity digital twins for simulating the impacts of connected and automated vehicles in microsimulation environments.

Populating agent-based models with agents who give rise to dynamics and scenarios of interest

Closed-loop autonomous-driving simulation with NVIDIA OmniDreams on AlpaSim: real-data rollouts with Alpamayo R1/1.5 drivers and counterfactual long-tail scenario generation (weather, lighting, injected 3D actors) via a generative world model.

Implemented stochastic CVaR model for the optimal asset allocation together with the Bootstrapping and the Monte Carlo scenario generation methods.

AI-assisted clinical simulation scenario generator for healthcare education.

Physics-informed, RL-aligned evaluation engine for autonomous driving: JAX/Flax NNX trajectory-diffusion world models on the Waymo Open Dataset, DPO and reward-guided adversarial steering toward the scenarios that break AV stacks, physics feasibility, occupancy flow, WOSAC-style realism metrics and sensor simulation.

基于 3DGS、Isaac Sim、CrowdES 与 Nav2 的室内动态人物场景生成、导航闭环评测与参数化扩增。

A simple program that allows players to randomize game scenarios for Stronghold Crusader HD. It features step-by-step rolling for 75 official maps, player counts, game modes, difficulty balance, and alliances. Designed with an immersive, game-stylized Java Swing UI and full dual-language support (EN/PL).

This project aims to synthesize realistic traffic scenarios from specifications in Temporal Logic. This work was done as part of my thesis @ TUM.

Test scenario generator for AD/ADAS using LLMS and scraped data

An AI-powered interactive storytelling platform that generates dynamic scenarios and narratives using local and cloud LLMs. Features scenario generation, memory management, and hopefully multi-provider LLM support.

YamlPrompt is a lightweight generative pipeline that enriches YAML files by replacing selected fields with AI‑generated text or images.

Deterministic local layer for permission, entitlement, route, and data-boundary sitting over Playwright, with Grassmann attention, and LLM-generated scenario-proposal workflows gated by a finite-state verifier.

Generate openscenarios using Z3 SMT solver.

Scripts and data used for the master's thesis "Portfolio selection with ES and regular vine copulae with EVT marginals" presented at the Wirtschaftsuniversität Wien in summer 2023. Code is structured for the purposes of the thesis and may not comply with best coding standards.

Natural language → CARLA scenarios, verified by swappable VLM judges. Closed-loop Generator / Verifier / Reflector.

Benchmark for scenario generation in two-stage stochastic and risk-averse supply chain design under climate shocks: copulas, scenario reduction, CVaR, VSS/EVPI, out-of-sample evaluation on a synthetic dairy network with known truth (Pyomo, HiGHS).

AI-native multi-agent simulator: Rust core + FastAPI service, synthetic RGB/depth/LiDAR/thermal, real MuJoCo physics, REST/SSE. Isaac Sim scaffold; SANA-WM not integrated.