Augmenting Differentiable Simulators with Neural Networks to Close the Sim2Real Gap
This work addresses the sim2real gap for robotics and simulation applications, presenting an incremental improvement by integrating neural networks into differentiable simulators.
The authors tackled the sim2real gap in articulated rigid-body dynamics by augmenting differentiable simulators with neural networks, achieving efficient parameter identification through gradient-based optimization and overcoming local optima with random search in experiments on real-world data and sim2sim transfers.
We present a differentiable simulation architecture for articulated rigid-body dynamics that enables the augmentation of analytical models with neural networks at any point of the computation. Through gradient-based optimization, identification of the simulation parameters and network weights is performed efficiently in preliminary experiments on a real-world dataset and in sim2sim transfer applications, while poor local optima are overcome through a random search approach.