Memory before Goal: Role of Pre-History Memory for Humanoid Loco-Manipulation
Memory before Goal: Role of Pre-History Memory for Humanoid Loco-Manipulation
This paper presents WholeBodyMem, an object-guided perceptual memory system that retains pre-history information for humanoid loco-manipulation in dynamic environments. By using SAM3 to track task objects and causally selecting key frames within an adaptive budget, the action expert integrates these frames via cross-attention with object tokens encoding appearance, geometry, and time. The authors also release HuMeGo, a benchmark featuring six simulated and six real-world Unitree G1 tasks. WholeBodyMem improves success rates from 29.1% to 72.6% in simulation and from 36.7% to 88.3% on the real robot, while also enhancing performance on standard SIMPLE tasks from 70.2% to 86.9%.
The authors introduce WholeBodyMem, an object-guided perceptual memory module that selects key frames from pre-history to support humanoid loco-manipulation in dynamic scenes, and HuMeGo, a benchmark comprising six simulated and six real-world Unitree G1 tasks; reported success rates of 72.6% in simulation and 88.3% on the real robot are attributed to the base policy augmented by WholeBodyMem.
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.
- paper: not_reported
Source excerpts and review record
Automatically extracted; no manual editorial approval recorded.
and uniform-memory baselines, raising its base policy's success from 29.1% to 72.6% in simulation and from 36.7% to 88.3% on the real robot, while also improving the standard SIMPLE tasks (70.2% to 86.9%).
Open source S4
Source：arXiv Robotics — research abstracts · arxiv.org