外观
论文arXiv 2505.122942025-05-18
PartDexTOG：语言驱动部件分析生成灵巧任务抓取
本文由 AI 根据公开内容整理，可能有误，请以原文为准：PartDexTOG: Generating Dexterous Task-Oriented Grasping via Language-driven Part Analysis（arXiv 2505.12294）
核心内容
针对灵巧手任务导向抓取难点，研究提出PartDexTOG框架，利用大语言模型对3D物体进行类别级与部件级功能描述生成，进而指导条件扩散模型产出对应部位的灵巧抓取姿态。该方法在OakInk-shape基准测试中全面超越现有方案，其中穿透体积降低3.58%、抓取位移减少2.87%、P-FID指标优化41.43%，验证了语义推理对几何生成的增强作用。
系统架构采用两阶段生成策略：首先由LLM基于自然语言任务指令输出结构化部件抓取描述，随后输入类别-部件条件扩散模型生成具体手部构型；为解决多候选组合的择优问题，设计了抓取-部件几何一致性度量机制。该方案将开放世界语义知识引入底层运动生成，使模型无需重新训练即可适配未见过的物体类别与操作任务，显著提升了具身智能在非结构化环境中的作业泛化能力。
关键看点
- 语义驱动生成：利用LLM生成类别与部件级抓取描述，作为条件引导扩散模型生成灵巧手姿态
- 性能指标突破：在OakInk-shape数据集上P-FID较SOTA提升41.43%，穿透体积与抓取位移分别优化3.58%和2.87%
- 几何一致性筛选：提出抓取与部件间的几何一致性度量，从多组生成结果中自动优选最合理组合
- 零样本泛化能力：借助LLM开放世界知识推理，支持对新物体类别与新操作任务的灵巧抓取生成
工程与产业观察
该工作验证了“语义理解+几何生成”范式在灵巧操作中的有效性，为具身智能提供了可复用的任务导向抓取技术路径。其核心启示在于：将高层语言推理与底层运动生成解耦，可通过语义先验缓解数据稀缺问题，对构建通用机器人操作技能库具有重要工程参考价值。
英文摘要 (Abstract)
Task-oriented grasping is a crucial yet challenging task in robotic manipulation. Despite the recent progress, few existing methods address task-oriented grasping with dexterous hands. Dexterous hands provide better precision and versatility, enabling robots to perform task-oriented grasping more effectively. In this paper, we argue that part analysis can enhance dexterous grasping by providing detailed information about the object's functionality. We propose PartDexTOG, a method that generates dexterous task-oriented grasps via language-driven part analysis. Taking a 3D object and a manipulation task represented by language as input, the method first generates the category-level and part-level grasp descriptions w.r.t the manipulation task by LLMs. Then, a category-part conditional diffusion model is developed to generate a dexterous grasp for each part, respectively, based on the generated descriptions. To select the most plausible combination of grasp and corresponding part from the generated ones, we propose a measure of geometric consistency between grasp and part. We show that our method greatly benefits from the open-world knowledge reasoning on object parts by LLMs, which naturally facilitates the learning of grasp generation on objects with different geometry and for different manipulation tasks. Our method ranks top on the OakInk-shape dataset over all previous methods, improving the Penetration Volume, the Grasp Displace, and the P-FID over the state-of-the-art by $3.58%$, $2.87%$, and $41.43%$, respectively. Notably, it demonstrates good generality in handling novel categories and tasks.
作者：Weishang Wu, Yifei Shi, Zhizhong Chen, Zhipong Cai
来源
本页事实对应的公开页面，以原文为准。