論文 Disruption-Resistant Deformable Object Manipulation on Basis of Online Shape Estimation and Prediction-Driven Trajectory Correction IEEE Robotics and Automation Letters,6(2):3809-3816 Apri( 20) Author:Daisuke Tanaka, Solvi Arnold, Kimitoshi Yamazaki Abstract:We consider the problem of deformable object manipulation with variable goal states and mid-manipulation disruptions. We propose an approach that integrates online shape estimation, prediction of shape transitions, and mid-manipulation trajectory correction. All functionalities are implemented using two neural network architectures. We apply this approach to the problem of cloth folding, and perform evaluation experiments in simulation and on robot hardware. We demonstrate that the system can achieve good approximation of given goal states, even when the manipulation process is disrupted by cloth slipping or external interference.
Fast and Flexible Multi-Step Cloth Manipulation Planning using an Encode-Manipulate-Decode NetworkNetwork (EM*D Net) Frontiers Neurorobotics,13 2019(May 31) Author:Solvi Arnold, Kimitoshi Yamazaki Abstract: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 <2.5 s for a three-step plan and is found to be robust to cloth self-occlusion, supporting the approach' viability for practical application.
An image recognition system aimed at search activities using cyber search and rescue dogs Journal of Field Robotics,36(4):677-695 2018(Dec. 03) Author:Solvi Arnold, Kazunori Ohno, Ryunosuke Hamada, Kimitoshi Yamazaki
二次学習の進化による心的表象の創発 人工知能学会誌,33(4):476-483 2018(Jul. 01) Author:有田隆也, Arnold Solvi, 鈴木麗璽
遠隔移動体を用いた災害時捜索活動のための画像認識システム 画像ラボ,29(2):7-14 2018(Feb.) Author:山崎公俊,Solvi Arnold
The emergence of executive functions by the evolution of second-order learning Artificial Life and Robotics,22(4):483–489 2017(Dec. 01) Author:Keisuke Daimon, Solvi Arnold, Reiji Suzuki, Takaya Arita
Unfolding of a rectangular cloth from unarranged starting shapes by a Dual-Armed robot with a mechanism for managing recognition error and uncertainty Advanced Robotics, Vol.31, Issue 10, pp. 544 – 556, http://dx.doi.org/10.1080/01691864.2017.1285722,31(10):544-556 2017(Feb. 09) Author:Hiroyuki Yuba, Solvi Arnold, Kimitoshi Yamazaki Keywords:Cloth unfolding, dual-armed manipulation, pinch and slide motion
Selection for Representation in Higher-Order Adaptation Minds & Machines,25:73-95 2015(Feb. 20) Author:Solvi Arnold, Reiji Suzuki, Takaya Arita
Why artificial intelligence cannot be creative without being Intelligent 11 Nagoya Journal of Philosophy,11 2014(Mar. 14) Author:Solvi Arnold
Selection for Reinforcement-Free Learning Ability as an Organizing Factor in the Evolution of Cognition Advances in Artificial Intelligence,2013 2013(Mar. 31) Author:Solvi Arnold, Reiji Suzuki, Takaya Arita
Visual Servoing Using Virtual Space for Both Learning and Task Execution The 2021 IEEE/SICE International Symposium on System Integration 2021(Jan. 13) Presenter:Tomoki Kawagoshi , Solvi Arnold , Kimitoshi Yamazaki
Picking of One Sheet of Cotton Cloth by Rolling up Using Cylindrical Brushes The 2021 IEEE/SICE International Symposium on System Integration 2021(Jan. 12) Presenter:Yuichi Kawasaki, Solvi Arnold , Kimitoshi Yamazaki
Learning Latent Representations of Environments and Grasping Motions for Fast Database Retrieval of Environments and Grasping Motions for Fast Database Retrieval of Collision-Free Motions 第38回日本ロボット学会学術講演会 RSJ2020 2020(Oct. 10) Presenter:Solvi Arnold, Keisuke Takeshita, Takashi Yamamoto, Rui Zhu, Kotaro Nagahama, Kimitoshi Yamazaki
Cloth Manipulation Planning with Mesh Representations and Incomplete Domain Knowledge 第38回日本ロボット学会学術講演会 RSJ2020 2020(Oct. 09) Presenter:Solvi Arnold, Daisuke Tanaka, Kimitoshi Yamazaki
Active Learning of the Cutting of Cooking Ingredients using Simulation with Object Splitting The 2020 IEEE/SICE International Symposium on System Integration 2020(Jan. 15) Presenter:Nahomi Ikegami, Solvi Arnold, Kotaro Nagahama, Kimitoshi Yamazaki
A Neural Network Approach for Fast and Flexible Multi-Step Cloth Manipulation Planning Third Machine Learning in Planning and Control of Robot Motion Workshop at ICRA2018 2018(May) Presenter:Solvi Arnold, Kimitoshi Yamazaki
Cloth Manipulation Planning by Back-propagation using a 3D Convolutional Auto-Encoder and aRecurrent Neural Network 第35回日本ロボット学会学術講演会 2017(Sep. 13) Presenter:Solvi Arnold, Kimitoshi Yamazaki
Hanging Work of T-Shirt in Consideration of Deformability and Stretchability IEEE International Conference on Information and Automation 2017(Jul. 18) Presenter:Yosuke Koishihara, Solvi Arnold, Kimitoshi Yamazaki, Takamitsu Matsubara
Real-time scene parsing by means of a convolutional neural network for mobile robots in disaster scenarios IEEE International Conference on Information and Automation 201-207 2017(Jul. 18) Presenter:Solvi Arnold, Kimitoshi Yamazaki Keywords:disaster robotics, rescue robots, computer vision, convolutional neural networks, scene parsing
Emergence of executive functions by evolved second-order learning The 22nd International Symposium on Artificial Life and Robotics (AROB2017) 2017(Jan. 18) Presenter:Keisuke Daimon, Solvi Arnold, Reiji Suzuki and Takaya Arita
Implicit Policies for Deformable Object Manipulation with Arbitrary Start and End States: A Novel Evolutionary Approach 2016 IEEE Conference on Robotics and Biomimetics 2016(Dec. 06) Presenter:Solvi Arnold, Kimitoshi Yamazaki
Unfolding of a Rectangular Cloth Based on Action Selection Depending on Recognition Uncertainty IEEE/SICE International Symposium on System Integration 623 – 628 2015(Dec.) Presenter:Hiroyuki Yuba, Solvi Arnold, Kimitoshi Yamazaki
Using Second Order Learning to Evolve Social Representation (Theory of Mind) The 8th International Conference on Bio-inspired Information and Communications Technologies 2014(Dec. 01) Presenter:Solvi Arnold, Reiji Suzuki, Takaya Arita
Evolution of Social Representation in Neural Networks Proceedings of the 12th European Conference on the Synthesis and Simulation of Living Systems (ECAL2013) 2013(Sep. 04) Presenter:Solvi Arnold, Reiji Suzuki, Takaya Arita
Evolution of Social Representation in Neural Networks 第27回人工知能学会全国大会 2013(Jun. 06) Presenter:Solvi Arnold, Reiji Suzuki, Takaya Arita
Second order learning and the evolution of mental representation Proceedings of the 13th International Conference on the Simulation and Synthesis of Living Systems (ALIFE2012) 2012(Jun. 21) Presenter:Solvi Arnold, Reiji Suzuki, Takaya Arita
Modelling Mental Representation as Evolved Second Order Learning Proceedings of the 17th International Symposium on Artificial Life and Robotics (AROB2012) 2012(Jan. 19) Presenter:Solvi Arnold, Reiji Suzuki, Takaya Arita
Neuro-cognitive Organization as a Side-effect of the Evolution of Learning Ability Proceedings of the 2011 IEEE Symposium on Artificial Life 2011(Apr. 11) Presenter:Solvi Arnold, Reiji Suzuki, Takaya Arita