GLST: Global-Local Spatio-Temporal Memory Decomposition Framework for Online Chained-Goal Navigation
GLST: Global-Local Spatio-Temporal Memory Decomposition Framework for Online Chained-Goal Navigation
This research proposes GLST, a memory-decomposition framework for online chained-goal navigation in partially observable environments. The method separates latent beliefs into global-local spatial and long-short temporal representations to support long-horizon reasoning. Goal-conditioned rewards are introduced to balance exploration and exploitation. The approach is validated through simulation and real-robot experiments using the Spot quadruped, demonstrating improved performance over existing baselines.
Source: arXiv Robotics — research abstracts · Read original article ↗
Loading article text…
What the source reports
Publisher-reported claims, with original evidence. These results have not been independently verified by RoboSignal.
- Environment
- SimulationOpen source S4
Source excerpts and review record
Automatically extracted; no manual editorial approval recorded.
mporal representations for long-horizon reasoning. We further introduce goal-conditioned rewards to balance exploration and exploitation under evolving goals. Extensive experiments in simulation across diverse terrains demonstrate improved chained-goal navigation performance over baselines. Real-robot experiments using the Spot quadruped further validate our memory decomposition framework.
Open source S4
Source：arXiv Robotics — research abstracts · arxiv.org