arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2305.04449cs.ROcs.AIcs.LG
收藏

DeformerNet：学习三维可变形物体的双臂操作

DeformerNet: Learning Bimanual Manipulation of 3D Deformable Objects

Kahlert School of Computing, University of Utah（犹他大学卡勒特计算学院）；NVIDIA Corporation（英伟达公司）

Bao Thach, Brian Y. Cho, Shing-Hei Ho, Tucker Hermans, Alan Kuntz

备注：
Submitted to IEEE Transactions on Robotics (T-RO). 20 pages, 27 figures. arXiv admin note: text overlap with arXiv:2110.04685
阅读 PDF
arXiv 页面 ↗HTML 原文 ↗

AI总结：

提出DeformerNet神经网络架构，通过点云学习三维可变形物体形状的低维表示，实现视觉伺服控制，在仿真和物理机器人上成功完成多种手术子任务并泛化至未知形状与刚度。

摘要：

从家庭护理到仓库分拣再到手术辅助等领域的应用，都需要机器人能够可靠地操作三维可变形物体的形状。弹性三维可变形物体的解析模型需要大量参数来描述决定物体形状时可能存在的无限自由度。以往尝试进行三维形状控制的方法依赖手工特征来表示物体形状，并且需要训练特定于物体的控制模型。我们通过使用新颖的 DeformerNet 神经网络架构克服了这些问题，该架构对被操作物体的局部视角点云和目标形状的点云进行操作，以学习物体形状的低维表示。这种形状嵌入使机器人能够学习一个视觉伺服控制器，该控制器计算期望的机器人末端执行器动作，以迭代方式将物体向目标形状变形。我们在仿真和物理机器人上证明，DeformerNet 能够可靠地泛化到训练期间未见过的物体形状和材料刚度，包括离体鸡肉肌肉组织。至关重要的是，使用 DeformerNet，机器人成功完成了三个手术子任务：牵拉（将组织移到一旁以进入其下方的部位）、组织包裹（如在主动脉支架放置等手术中的子任务）以及连接两块管状组织（吻合术中的子任务）。

英文摘要：

Applications in fields ranging from home care to warehouse fulfillment to surgical assistance require robots to reliably manipulate the shape of 3D deformable objects. Analytic models of elastic, 3D deformable objects require numerous parameters to describe the potentially infinite degrees of freedom present in determining the object's shape. Previous attempts at performing 3D shape control rely on hand-crafted features to represent the object shape and require training of object-specific control models. We overcome these issues through the use of our novel DeformerNet neural network architecture, which operates on a partial-view point cloud of the manipulated object and a point cloud of the goal shape to learn a low-dimensional representation of the object shape. This shape embedding enables the robot to learn a visual servo controller that computes the desired robot end-effector action to iteratively deform the object toward the target shape. We demonstrate both in simulation and on a physical robot that DeformerNet reliably generalizes to object shapes and material stiffness not seen during training, including ex vivo chicken muscle tissue. Crucially, using DeformerNet, the robot successfully accomplishes three surgical sub-tasks: retraction (moving tissue aside to access a site underneath it), tissue wrapping (a sub-task in procedures like aortic stent placements), and connecting two tubular pieces of tissue (a sub-task in anastomosis).

机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。

↑