多尺度对比学习提升机器人抓取适应性
MISCGrasp: Leveraging Multiple Integrated Scales and Contrastive Learning for Enhanced Volumetric Grasping
- 通过双注意力机制融合高低层特征，兼顾细节与整体结构
- 多尺度对比学习使不同尺度特征保持一致性，提升泛化能力
- 在模拟与真实场景中均优于基线方法，适合复杂桌面抓取
机器人抓取面临应对不同形状和尺寸物体的挑战。本文提出MISCGrasp，一种结合多尺度特征提取与对比特征增强的体素抓取方法。通过洞察变压器实现高层与低层特征间的查询式交互，而赋能变压器则选择性关注最顶层特征，协同平衡对细微几何细节与整体几何结构的关注。此外，MISCGrasp利用多尺度对比学习，挖掘正样本间的相似性，确保多尺度特征的一致性。大量模拟与真实环境实验表明，MISCGrasp在桌面清理工任务中优于基线及变体方法。
原文摘要 · Abstract (English)
Robotic grasping faces challenges in adapting to objects with varying shapes and sizes. In this paper, we introduce MISCGrasp, a volumetric grasping method that integrates multi-scale feature extraction with contrastive feature enhancement for self-adaptive grasping. We propose a query-based interaction between high-level and low-level features through the Insight Transformer, while the Empower Transformer selectively attends to the highest-level features, which synergistically strikes a balance between focusing on fine geometric details and overall geometric structures. Furthermore, MISCGrasp utilizes multi-scale contrastive learning to exploit similarities among positive grasp samples, ensuring consistency across multi-scale features. Extensive experiments in both simulated and real-world environments demonstrate that MISCGrasp outperforms baseline and variant methods in tabletop decluttering tasks. More details are available at https://miscgrasp.github.io/.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。