Learning Closed-Loop Dough Manipulation Using a Differentiable Reset Module

نویسندگان

چکیده

Deformable object manipulation has many applications such as cooking and laundry folding in our daily lives. Manipulating elastoplastic objects dough is particularly challenging because lacks a compact state representation requires contact-rich interactions. We consider the task of flattening piece into specific shape from RGB-D images. While seemingly intuitive for humans, there exist local optima common approaches naive trajectory optimization. propose novel optimizer that optimizes through differentiable “reset” module, transforming single-stage, fixed-initialization multistage, multi-initialization where all stages are optimized jointly. then train closed-loop policy on demonstrations generated by optimizer. Our receives partial point clouds input, allowing ease transfer simulation to real world. show can perform real-world manipulation, ball target shape.

برای دانلود باید عضویت طلایی داشته باشید

برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

Automated Droplet Manipulation Using Closed-Loop Axisymmetric Drop Shape Analysis.

Droplet manipulation plays an important role in a wide range of scientific and industrial applications, such as synthesis of thin-film materials, control of interfacial reactions, and operation of digital microfluidics. Compared to micron-sized droplets, which are commonly considered as spherical beads, millimeter-sized droplets are generally deformable by gravity, thus introducing nonlinearity...

متن کامل

Closed-Loop Object Recognition Using Reinforcement Learning

Current computer vision systems whose basic methodology is open-loop or filter type typically use image segmentation followed by object recognition algorithms. These systems are not robust for most real-world applications. In contrast, the system presented here achieves robust performance by using reinforcement learning to induce a mapping from input images to corresponding segmentation paramet...

متن کامل

Closed-loop Visual Grasping and Manipulation

Sensors play an important role in providing feedback to robotic control systems. Vision can be an eeective sensing modality due to its speed, low cost, and ex-ibility. It can also serve as an external sensor that can provide control information for devices that lack internal sensing. Many robot grippers do not have internal sensors. These grippers can be prone to error as they operate under ope...

متن کامل

Distance protection closed-loop testing using RTDS

This paper presents a distance protection test procedure by applying the Real-Time Digital Simulator (RTDS) of a power system. RTDS is a tool to design, develop, and test power-system protection. The RTDS enables real-time computation of electromagnetic phenomena with a calculation time step of even 50μs. The hardware allows the import and export of many signals from the simulator to an ext...

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

ژورنال

عنوان ژورنال: IEEE robotics and automation letters

سال: 2022

ISSN: ['2377-3766']

DOI: https://doi.org/10.1109/lra.2022.3191239