DiffGen: Robot Demonstration Generation via Differentiable Physics Simulation, Differentiable Rendering, and Vision-Language Model

TL;DR

DiffGen is a framework that combines differentiable physics, rendering, and vision-language models to automatically generate realistic robot demonstrations from natural language instructions, improving efficiency and reducing manual effort.

Contribution

It introduces a novel integration of differentiable simulation, rendering, and vision-language models for automatic robot demonstration generation from natural language.

Findings

Efficient generation of robot demonstrations with minimal human effort.

Effective alignment of language instructions with simulated observations.

Demonstrated success in various manipulation scenarios.

Abstract

Generating robot demonstrations through simulation is widely recognized as an effective way to scale up robot data. Previous work often trained reinforcement learning agents to generate expert policies, but this approach lacks sample efficiency. Recently, a line of work has attempted to generate robot demonstrations via differentiable simulation, which is promising but heavily relies on reward design, a labor-intensive process. In this paper, we propose DiffGen, a novel framework that integrates differentiable physics simulation, differentiable rendering, and a vision-language model to enable automatic and efficient generation of robot demonstrations. Given a simulated robot manipulation scenario and a natural language instruction, DiffGen can generate realistic robot demonstrations by minimizing the distance between the embedding of the language instruction and the embedding of the simulated observation after manipulation. The embeddings are obtained from the vision-language model, and the optimization is achieved by calculating and descending gradients through the differentiable simulation, differentiable rendering, and vision-language model components, thereby accomplishing the specified task. Experiments demonstrate that with DiffGen, we could efficiently and effectively generate robot data with minimal human effort or training time.