Accelerating Visual-Policy Learning through Parallel Differentiable Simulation

Haoxiang You, Yilang Liu, Ian Abraham

Keywords: visual policy learning, analytical policy gradient, differentiable simulation

TL;DR: We introduce a computationally efficient method for visual policy learning using differentiable simulation.

Abstract:

In this work, we propose a computationally efficient algorithm for visual policy learning that leverages differentiable simulation and first-order analytical policy gradients.
Our approach decouple the rendering process from the computation graph, enabling seamless integration with existing differentiable simulation ecosystems without the need for specialized differentiable rendering software.
This decoupling not only reduces computational and memory overhead but also effectively attenuates the policy gradient norm, leading to more stable and smoother optimization.
We evaluate our method on standard visual control benchmarks using modern GPU-accelerated simulation.
Experiments show that our approach significantly reduces wall-clock training time and consistently outperforms all baseline methods in terms of final returns.
Notably, on complex tasks such as humanoid locomotion, our method achieves a $4\times$ improvement in final return, and successfully learns a humanoid running policy within 4 hours on a single GPU.
Videos and code are available on https://haoxiangyou.github.io/Dva_website

Primary Area: Reinforcement learning (e.g., decision and control, planning, hierarchical RL, robotics)

Submission Number: 5262