Value-Preserving Action Quantization
Value-Preserving Action Quantization
This research proposes VAQ, a method for action quantization that preserves state reachability, enhancing data efficiency and robustness in robotic policy learning. When combined with MCTS, VAQ-MCTS achieves performance comparable to state-of-the-art methods while enabling zero-shot adaptation to novel environments, as demonstrated through hardware experiments.
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.
Reported numbers
object pushing trials
30
demonstration data
10 minutes
Source excerpts and review record
Automatically extracted; no manual editorial approval recorded.
-the-art behavioral cloning methods. VAQ-MCTS matches or exceeds their performance under standard settings, while substantially improving data efficiency and robustness to heterogeneous demonstrations, while enabling zero-shot adaptation to novel environments. Hardware experiments further demonstrate the effectiveness of VAQ-MCTS, completing all 30 real-world object pushing trials with only 10 minutes of demonstratio
Open source S5
Source：arXiv Robotics — research abstracts · arxiv.org