V-Dreamer: Automating Robotic Simulation and Trajectory Synthesis via Video Generation Priors

TL;DR

V-Dreamer introduces an automated framework that generates diverse, simulation-ready manipulation environments and trajectories from natural language, enabling scalable robot training and effective sim-to-real transfer.

Contribution

The paper presents a novel generative pipeline combining language models, 3D generative models, and visual priors to create diverse, physically grounded simulation environments and trajectories from natural language instructions.

Findings

Policies trained on synthesized data generalize to unseen objects.

The approach achieves effective sim-to-real transfer in manipulation tasks.

Generated environments support high visual diversity and physical fidelity.

Abstract

Training generalist robots demands large-scale, diverse manipulation data, yet real-world collection is prohibitively expensive, and existing simulators are often constrained by fixed asset libraries and manual heuristics. To bridge this gap, we present V-Dreamer, a fully automated framework that generates open-vocabulary, simulation-ready manipulation environments and executable expert trajectories directly from natural language instructions. V-Dreamer employs a novel generative pipeline that constructs physically grounded 3D scenes using large language models and 3D generative models, validated by geometric constraints to ensure stable, collision-free layouts. Crucially, for behavior synthesis, we leverage video generation models as rich motion priors. These visual predictions are then mapped into executable robot trajectories via a robust Sim-to-Gen visual-kinematic alignment module utilizing CoTracker3 and VGGT. This pipeline supports high visual diversity and physical fidelity without manual intervention. To evaluate the generated data, we train imitation learning policies on synthesized trajectories encompassing diverse object and environment variations. Extensive evaluations on tabletop manipulation tasks using the Piper robotic arm demonstrate that our policies robustly generalize to unseen objects in simulation and achieve effective sim-to-real transfer, successfully manipulating novel real-world objects.