Fast and Flexible Multi-Step Cloth Manipulation Planning Using an Encode-Manipulate-Decode Network (EM*D Net)
Fast and Flexible Multi-Step Cloth Manipulation Planning Using an Encode-Manipulate-Decode Network (EM*D Net)
复制标题
DOI：
10.3389/fnbot.2019.00022
复制
发表时间：
2019-05-31
影响因子：
3.1
通讯作者：
Yamazaki, Kimitoshi
中科院分区：
文献类型：
--
作者：
Arnold, Solvi;Yamazaki, Kimitoshi
我们提出了一种深度神经网络架构，编码-操纵-解码（EM*D）网络，用于可变形物体的快速操纵规划。我们证明了它的有效性模拟布。该网络由3D卷积编码器和解码器模块组成，这些模块将布料状态映射到潜在空间，并从潜在空间映射，中间有一个“操纵模块”，它学习布料动态的前向模型。在潜在空间中的操作系统。操纵模块的架构专门用于其作为前向模型的角色，通过在每一层的剩余连接和重复输入来迭代地修改状态表示。我们训练网络从预操纵布料状态和操纵输入预测操纵后布料状态。通过端到端训练网络，我们迫使编码器和解码器模块学习一个潜在的状态表示，以便于操作模块进行修改。我们表明，该网络可以从6,000个操作示例的训练数据集中实现良好的泛化。比较实验没有架构专业化的操作模块显示性能降低，确认我们的架构的好处。通过执行误差反向传播w.r.t.操纵输入。在规划期间反复使用操纵网络允许生成多步规划。我们展示了多达三个操作的计划的结果，一般表现出良好的近似的目标状态。计划生成需要
We propose a deep neural network architecture, the Encode-Manipulate-Decode (EM*D) net, for rapid manipulation planning on deformable objects. We demonstrate its effectiveness on simulated cloth. The net consists of 3D convolutional encoder and decoder modules that map cloth states to and from latent space, with a "manipulation module" in between that learns a forward model of the cloth's dynamics w.r.t. the manipulation repertoire, in latent space. The manipulation module's architecture is specialized for its role as a forward model, iteratively modifying a state representation by means of residual connections and repeated input at every layer. We train the network to predict the post-manipulation cloth state from a pre-manipulation cloth state and a manipulation input. By training the network end-to-end, we force the encoder and decoder modules to learn a latent state representation that facilitates modification by the manipulation module. We show that the network can achieve good generalization from a training dataset of 6,000 manipulation examples. Comparative experiments without the architectural specializations of the manipulation module show reduced performance, confirming the benefits of our architecture. Manipulation plans are generated by performing error back-propagation w.r.t. the manipulation inputs. Recurrent use of the manipulation network during planning allows for generation of multi-step plans. We show results for plans of up to three manipulations, demonstrating generally good approximation of the goal state. Plan generation takes