LangDriveCTRL is a natural-language-controllable framework for editing real-world driving videos to synthesize diverse traffic scenarios. It represents each video as an explicit 3D scene graph, decomposing the scene into a static background and dynamic object nodes. To enable fine-grained editing and realism, it introduces a feedback-driven agentic pipeline. An Orchestrator converts user instructions into executable graphs that coordinate specialized multi-modal agents and tools. An Object Grounding Agent aligns free-form text with target object nodes in the scene graph; a Behavior Editing Agent generates multi-object trajectories from language instructions; and a Behavior Reviewer Agent iteratively reviews and refines the generated trajectories. The edited scene graph is rendered and harmonized using a video diffusion tool, and then further refined by a Video Reviewer Agent to ensure photorealism and appearance alignment. LangDriveCTRL supports both object node editing (removal, insertion, and replacement) and multi-object behavior editing from natural-language instructions. Quantitatively, it achieves nearly 2× higher instruction alignment than the previous SoTA, with superior photorealism, structural preservation, and traffic realism. Project page is available at: https://yunhe24.github.io/langdrivectrl/.
@article{arxiv.2512.17445,
title = {LangDriveCTRL: Natural Language Controllable Driving Scene Editing with Multi-modal Agents},
author = {Yun He and Francesco Pittaluga and Ziyu Jiang and Matthias Zwicker and Manmohan Chandraker and Zaid Tasneem},
journal= {arXiv preprint arXiv:2512.17445},
year = {2026}
}