外观
论文arXiv 2609.215721 个月前
SABER：学习用于腿式运动的基于注意力的语义可供性
SABER针对腿式机器人不当接触风险，用地形可供性图和语义偏置注意力在Unitree B2上验证。
宇树 B2原文链接：SABER: Learning Attention-based Semantic Affordance for Legged Locomotion（arXiv 2609.21572 · 发布日期：2026-09-18）
SABER: Learning Attention-based Semantic Affordance for Legged Locomotion
SABER针对腿式机器人不当接触风险，用地形可供性图和语义偏置注意力在Unitree B2上验证。
核心要点
- 仅利用地形几何会忽略语义含义，几何可通行区域也可能不适合接触。
- 该方法是无规划器强化学习策略，联合推理地形几何与语义接触许可。
- 策略使用统一地形可供性图，并以接触代价门控的有符号语义偏置增强交叉注意力。
- 消融显示去除语义偏置使禁区接触增加55%，速度跟踪不变；真机展示跨环境迁移。
验证范围与局限：验证范围为Unitree B2真机、室内外环境和四类语义障碍。
英文摘要 (Abstract)
Perceptive legged locomotion has advanced rapidly by integrating terrain geometry into learned policies, yet the integration of terrain meaning remains sparse: a pipe, a patch of grass, or a fragile box may be geometrically traversable while being inappropriate for contact. In industrial environments, where legged robots increasingly operate, a single misplaced step can damage fragile equipment, destabilize the robot, or endanger the site. To address this, we introduce SABER, a planner-free reinforcement-learning policy that jointly reasons about terrain geometry and semantic contact permission. The policy consumes a unified terrain-affordance map, where each cell encodes local 3D geometry and a semantic contact cost. We augment cross-attention with a learned, signed semantic bias: an additive term on the attention logits, gated by the contact cost, that reweights flagged cells by their distance from the nearest foot. A hazard therefore reshapes attention where it can still affect the next foothold, and its influence fades where it cannot. The resulting policy selects footholds on permitted support and keeps the leg clear of forbidden regions throughout the swing phase. We perform a systematic ablation that isolates the contribution of each architectural component; removing the semantic bias alone increases forbidden contacts by 55% while velocity tracking is unchanged. We validate the policy on a Unitree B2, demonstrating sim-to-real semantic contact selection across indoor and outdoor environments and four semantic obstacle classes.
作者：Hari P. Palanivelu, Samuel Sze, Kennard Garrison Johannes, Albertus Hendrawan Adiwahono, Meng Yee Chuah
来源
本页事实对应的公开页面，以原文为准。