Composite Gaussian Processes Flows for Learning Discontinuous Multimodal Policies
This addresses robotic policy learning for real-world tasks, but it appears incremental as it combines existing methods like OMGPs and CNFs.
The paper tackled learning control policies for robotic tasks with challenges like multimodality and discontinuities by proposing Composite Gaussian Processes Flows (CGP-Flows), which improved success rates in simulations compared to baselines, with statistically significant differences in chi-square tests.
Learning control policies for real-world robotic tasks often involve challenges such as multimodality, local discontinuities, and the need for computational efficiency. These challenges arise from the complexity of robotic environments, where multiple solutions may coexist. To address these issues, we propose Composite Gaussian Processes Flows (CGP-Flows), a novel semi-parametric model for robotic policy. CGP-Flows integrate Overlapping Mixtures of Gaussian Processes (OMGPs) with the Continuous Normalizing Flows (CNFs), enabling them to model complex policies addressing multimodality and local discontinuities. This hybrid approach retains the computational efficiency of OMGPs while incorporating the flexibility of CNFs. Experiments conducted in both simulated and real-world robotic tasks demonstrate that CGP-flows significantly improve performance in modeling control policies. In a simulation task, we confirmed that CGP-Flows had a higher success rate compared to the baseline method, and the success rate of GCP-Flow was significantly different from the success rate of other baselines in chi-square tests.