QuadKAN: KAN-Enhanced Quadruped Motion Control via End-to-End Reinforcement Learning

TL;DR

QuadKAN introduces a spline-parameterized reinforcement learning framework that effectively integrates vision and proprioception for robust quadruped locomotion across diverse terrains, improving efficiency and safety.

Contribution

The paper presents QuadKAN, a novel spline-parameterized policy using KANs, enhancing sample efficiency, interpretability, and robustness in vision-guided quadruped control.

Findings

Outperforms SOTA in diverse terrains and obstacle scenarios.

Reduces action jitter and energy consumption.

Provides interpretable posture-action sensitivities.

Abstract

We address vision-guided quadruped motion control with reinforcement learning (RL) and highlight the necessity of combining proprioception with vision for robust control. We propose QuadKAN, a spline-parameterized cross-modal policy instantiated with Kolmogorov-Arnold Networks (KANs). The framework incorporates a spline encoder for proprioception and a spline fusion head for proprioception-vision inputs. This structured function class aligns the state-to-action mapping with the piecewise-smooth nature of gait, improving sample efficiency, reducing action jitter and energy consumption, and providing interpretable posture-action sensitivities. We adopt Multi-Modal Delay Randomization (MMDR) and perform end-to-end training with Proximal Policy Optimization (PPO). Evaluations across diverse terrains, including both even and uneven surfaces and scenarios with static or dynamic obstacles, demonstrate that QuadKAN achieves consistently higher returns, greater distances, and fewer collisions than state-of-the-art (SOTA) baselines. These results show that spline-parameterized policies offer a simple, effective, and interpretable alternative for robust vision-guided locomotion. A repository will be made available upon acceptance.