GLST: Global-Local Spatio-Temporal Memory Decomposition Framework for Online Chained-Goal Navigation
We propose Global-Local Spatio-Temporal Belief Learning (GLST), a memory-decomposition framework that separates latent beliefs into global-local spatial and long-short temporal representations for long-horizon reasoning.
ProofPaper ↗
Key points
- Autonomous navigation in real-world environments often requires long-horizon operation under partial observability and evolving goals.
- We study online chained-goal navigation, where goals evolve through existing-goal redirection or new-goal emergence without explicit transition signals.
- 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.
Sources (1)
- [1]GLST: Global-Local Spatio-Temporal Memory Decomposition Framework for Online Chained-Goal NavigationarXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 13, 08:15 AM
We propose Global-Local Spatio-Temporal Belief Learning (GLST), a memory-decomposition framework that separates latent beliefs into global-local spatial and long-short temporal representations for long-horizon reasoning.
Autonomous navigation in real-world environments often requires long-horizon operation under partial observability and evolving goals.
Extractive summary: sentences quoted from the sources.