Human-like continual robot learning based on three-level computational energy cost regulation

基于三级计算能量成本调节的类人持续学习机器人

• 批准号: 22H03670
• 负责人: OZTOP Erhan
• 金额: $ 9.57万
• 依托单位: Osaka University
• 依托单位国家: 日本
• 项目类别: Grant-in-Aid for Scientific Research (B)
• 财政年份: 2022
• 资助国家: 日本
• 起止时间: 2022-04-01 至 2025-03-31
• 项目状态: 未结题
• 来源: https://kaken.nii.ac.jp/en/grant/KAKENHI-PROJECT-22H03670/
• 关键词: 計算エネルギーコスト; 継続的な行動学習; 概念形成; 内発的動機; スキル移転; Lifelong Robot Learning; Learning Progress; Knowledge Transfer; Multitask Learning; Symbol Formation

In the first year of the project an appropriate robot simulator environment is selected (Pybullet) and the software platform for Lifelong Robot Learning (LRL) model has been developed on it. A robotic arm with three tasks (T1,T2,T3) is considered for LRL. The robot action is modeled as hitting objects with different angles. LRL tasks are set as the prediction of the effects of the actions in the three different environments, free space (T1), wall with changing orientation (T2) , L-shaped wall with changing orientation (T3). Task execution is based on Learning Progress (LP) whereas ‘neural cost’ consideration is left for next year. A basic knowledge transfer architecture is developed among the neural networks of each task. The symbol formation component is also explored but not incorporated into the simulated LRL model. Parallel to the development of the LRL model, supporting work is conducted and several publications are produced, and a workshop in IROS 2022 is held together with collaborators. In one line of research, work on symbol formation by the use of discrete units in the latent layers of deep neural networks is studied [2]. In addition, a work on robotic trust is conducted with collaborators which uses ‘neural computational cost’ for forming trust in social partners [3]. Therefore the LRL model can be extended to include trust formation, even though it was not directly part of the initial proposal. In addition, for supporting human-robot related tasks some work is devoted to teaching robots how to correct errors based on human demonstration.

在项目的第一年中，选择了适当的机器人模拟器环境（PYBULLET），并在其上开发了终身机器人学习（LRL）模型的软件平台。对于LRL，考虑了一个具有三个任务（T1，T2，T3）的机器人臂（T1，T2，T3）。机器人动作被建模为击中不同角度的对象。 LRL任务被设置为对三种不同环境中动作影响的预测，即自由空间（T1），带有变化方向的墙（T2），L形墙，带有变化的方向（T3）。任务执行是基于学习进度（LP）的，而“神经成本”考虑明年。在每个任务的神经元网络中开发了基本知识转移体系结构。还探索了符号形成组件，但未纳入模拟LRL模型。与LRL模型的开发平行，进行了支持工作，并制作了几个出版物，并与合作者一起举行了IROS 2022的研讨会。在一条研究中，研究了通过在深神经网络的潜在层中使用离散单元来进行符号形成的工作[2]。此外，与合作者一起进行了关于机器人信任的工作，该协作使用“神经计算成本”来形成社会伙伴的信任[3]。因此，即使它不是初始提案的直接一部分，LRL模型也可以扩展到包括信任形成。此外，为了支持与人类机器人相关的任务，某些工作专门用于教机器人如何根据人类示威来纠正错误。

项目成果

Trust in robot-robot scaffolding

对机器人脚手架的信任

• DOI: 10.1109/tcds.2023.3235974
• 发表时间: 2023
• 期刊: IEEE Transactions on Cognitive and Developmental Systems
• 影响因子: 5
• 作者: Kirtay Murat;Hafner Verena V.;Asada Minoru;Oztop Erhan
• 通讯作者: Oztop Erhan

Multimodal reinforcement learning for partner specific adaptation in robot-multi-robot interaction

机器人与多机器人交互中伙伴特定适应的多模态强化学习

• DOI: 10.1109/humanoids53995.2022.10000205
• 发表时间: 2022
• 期刊: IEEE Proceedings on Humanoids 2022, Ginowan, Japan
• 作者: Kirtay Murat;Hafner Verena V.;Asada Minoru;Kuhlen Anna K.;Oztop Erhan
• 通讯作者: Oztop Erhan

Bimanual Rope Manipulation Skill Synthesis through Context Dependent Correction Policy Learning from Human Demonstration

• DOI: 10.1109/icra48891.2023.10160895
• 发表时间: 2022-09
• 期刊: 2023 IEEE International Conference on Robotics and Automation (ICRA)
• 作者: T. Akbulut;G. Girgin;A. Mehrabi;M. Asada;Emre Ugur;E. Oztop

Bogazici University/Ozyegin University(トルコ)

海峡大学/奥济耶金大学（土耳其）

DeepSym: Deep Symbol Generation and Rule Learning for Planning from Unsupervised Robot Interaction

DeepSym：用于无监督机器人交互规划的深度符号生成和规则学习

• DOI: 10.1613/jair.1.13754
• 发表时间: 2022
• 期刊: Journal of Artificial Intelligence Research
• 影响因子: 5
• 作者: Ahmetoglu Alper;Seker M. Yunus;Piater Justus;Oztop Erhan;Ugur Emre
• 通讯作者: Ugur Emre