Article
Computer Science, Artificial Intelligence
Changlin Han, Zhiyong Peng, Yadong Liu, Jingsheng Tang, Yang Yu, Zongtan Zhou
Summary: In this paper, an overfitting-avoiding goal-guided exploration method (OGE) is proposed. It generates auxiliary goals following the Wasserstein-distance-based optimal transport geodesic and has a generation region that accounts for agent generalizability. Our method outperforms state-of-the-art methods in hard-exploration multi-goal robotic manipulation tasks, achieving high learning efficiency and successfully guiding the agent to achieve hard goals.
Article
Chemistry, Multidisciplinary
Nikolai V. Krivoshchapov, Michael G. Medvedev
Summary: Inverse kinematics provides a mathematically strict solution for handling cyclic molecules in conformational searches, allowing for the narrowing of conformational space and improving the accuracy and efficiency of models.
WILEY INTERDISCIPLINARY REVIEWS-COMPUTATIONAL MOLECULAR SCIENCE
(2023)
Article
Psychology, Multidisciplinary
Jason Kostrna
Summary: This study examined the effects of time constraints on anxiety, attention, performance, and mechanics of basketball free-throw shooting. The results showed that participants performed better in untimed conditions compared to timed conditions, and the goal-oriented condition had similar performance to the untimed condition. Additionally, joint consistency in the elbow and knee increased in the untimed condition.
FRONTIERS IN PSYCHOLOGY
(2022)
Article
Computer Science, Artificial Intelligence
Rania Rayyes, Heiko Donat, Jochen Steil
Summary: This article introduces the challenge of high sample complexity in online robot learning and proposes new methods to overcome this problem. The author utilizes exploration and intrinsic motivation signals to drive robot online learning and introduces an episodic online mental replay method to accelerate the learning process. The efficiency and applicability of these methods are demonstrated with experiments on a physical robot.
IEEE TRANSACTIONS ON COGNITIVE AND DEVELOPMENTAL SYSTEMS
(2022)
Article
Computer Science, Artificial Intelligence
Filipe Gama, Maksym Shcherban, Matthias Rolf, Matej Hoffmann
Summary: Early integration of tactile sensing into motor coordination is the norm in animals, but still a challenge for robots. Tactile exploration through touches on the body gives rise to first body models and bootstraps further development such as reaching competence. Reaching to one's own body requires connections of the tactile and motor space only. Still, the problems of high dimensionality and motor redundancy persist.
IEEE TRANSACTIONS ON COGNITIVE AND DEVELOPMENTAL SYSTEMS
(2023)
Article
Automation & Control Systems
Sebastien Forestier, Remy Portelas, Yoan Mollard, Pierre-Yves Oudeyer
Summary: Intrinsically motivated spontaneous exploration is crucial for autonomous developmental learning in human children, and the Intrinsically Motivated Goal Exploration Processes (IMGEP) algorithmic approach enables similar autonomous learning properties in machines. The IMGEP architecture relies on principles such as self-generation of goals, goal selection based on intrinsic rewards, and systematic reuse of acquired information. The AMB, a highly efficient form of IMGEP, can automatically generate a learning curriculum and has been demonstrated in various experimental setups.
JOURNAL OF MACHINE LEARNING RESEARCH
(2022)
Article
Computer Science, Artificial Intelligence
Adrien Laversanne-Finot, Alexandre Pere, Pierre-Yves Oudeyer
Summary: Finding algorithms that enable agents to efficiently and autonomously discover a wide variety of skills remains a challenge in Artificial Intelligence. Using Intrinsically Motivated Goal Exploration Processes (IMGEPs) and deep representation learning algorithms can effectively help agents explore complex environments and reduce the burden of designing goal spaces.
FRONTIERS IN NEUROROBOTICS
(2021)
Article
Robotics
Kei Kase, Ai Tateishi, Tetsuya Ogata
Summary: This study focuses on the concept of pseudo-rehearsal and proposes a framework that can be jointly trained with task trajectories and rehearsed motor babbling trajectories. It allows robots to retain motor skills acquired from motor babbling and exhibit improved performance in task execution.
IEEE ROBOTICS AND AUTOMATION LETTERS
(2022)
Article
Robotics
Reinis Cimurs, Il Hong Suh, Jin Han Lee
Summary: This letter presents an autonomous navigation system for exploring unknown environments using deep reinforcement learning (DRL). The system selects optimal waypoints and learns a motion policy for local navigation to guide the robot towards a global goal. Experimental results show that the proposed method outperforms similar exploration methods in complex static and dynamic environments.
IEEE ROBOTICS AND AUTOMATION LETTERS
(2022)
Article
Robotics
Gian Maria Marconi, Raffaello Camoriano, Lorenzo Rosasco, Carlo Ciliberto
Summary: With recent advances in machine learning, a structured prediction algorithm that combines data-driven strategies with the model provided by a forward kinematics function has been proposed. This approach effectively solves the inverse kinematics problem of a redundant robot arm by ensuring predicted joint configurations are well within the robot's constraints.
IEEE ROBOTICS AND AUTOMATION LETTERS
(2021)
Article
Automation & Control Systems
M. N. Vu, F. Beck, M. Schwegel, C. Hartl-Nesic, A. Nguyen, A. Kugi
Summary: Solving the analytical inverse kinematics (IK) of redundant manipulators in real time is a challenging problem. This paper presents a real-time framework that parameterizes the analytical IK of the redundant manipulator using redundancy parameters, combined with a target pose to yield a unique IK solution. Unlike existing approaches, the proposed framework directly learns these parameters using a neural network, providing the optimal IK solution with respect to manipulability and closeness to the current robot configuration.
Article
Computer Science, Artificial Intelligence
Rania Rayyes, Heiko Donat, Jochen Steil, Michael Spranger
Summary: This article proposes a novel extrinsic-intrinsic motivation learning scheme to accelerate learning by combining intrinsic motivation with learning from observation. The scheme includes four elements: 1) a probabilistic intrinsic motivation signal to spark the robot's interest; 2) a probabilistic extrinsic motivation signal to expand the robot's knowledge through learning from observation; 3) novelty detection; and 4) novelty degree methods for the robot to autonomously decide how and when to explore.
IEEE TRANSACTIONS ON COGNITIVE AND DEVELOPMENTAL SYSTEMS
(2023)
Article
Automation & Control Systems
Daniel Cagigas-Muniz
Summary: The inverse kinematics problem in articulated robots refers to obtaining joint rotation angles using the robot end effector position and orientation tool. This study proposes and analyzes different techniques involving artificial neural networks (ANNs) to solve this problem. The results demonstrate that the proposed original bootstrap sampling and hybrid methods can greatly improve the performance of approaches using only one ANN. While these improvements do not completely solve the inverse kinematics problem in articulated robots, they lay the foundations for designing and developing more effective and efficient controllers.
ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE
(2023)
Proceedings Paper
Computer Science, Artificial Intelligence
Rohan Chitnis, Tom Silver, Joshua Tenenbaum, Leslie Pack Kaelbling, Tomas Lozano-Perez
Summary: The study introduces a goal-literal babbling (GLIB) method inspired by human curiosity for efficient exploration in relational model-based reinforcement learning, with experimental results showing superior performance in various tasks.
THIRTY-FIFTH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE, THIRTY-THIRD CONFERENCE ON INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE AND THE ELEVENTH SYMPOSIUM ON EDUCATIONAL ADVANCES IN ARTIFICIAL INTELLIGENCE
(2021)
Article
Psychology, Multidisciplinary
Adam J. Toth, Fazilat Hojaji, Mark J. Campbell
Summary: This study aims to use existing kinematic data to investigate whether there are differences in specific phases of target acquisition movements between gamers of different expertise levels. The results show that gamers with higher expertise demonstrate superior motor planning and sensory-motor integration, which can be further improved through training.
COMPUTERS IN HUMAN BEHAVIOR
(2023)
Article
Robotics
M. Rolf, K. Neumann, J. F. Queisser, R. F. Reinhart, A. Nordmann, J. J. Steil
Article
Computer Science, Artificial Intelligence
Klaus Neumann, Matthias Rolf, Jochen Jakob Steil
INTERNATIONAL JOURNAL OF UNCERTAINTY FUZZINESS AND KNOWLEDGE-BASED SYSTEMS
(2013)
Article
Computer Science, Artificial Intelligence
Matthias Rolf, Jochen J. Steil
Article
Computer Science, Artificial Intelligence
Matthias Rolf, Marc Hanheide, Katharina J. Rohlfing
IEEE TRANSACTIONS ON AUTONOMOUS MENTAL DEVELOPMENT
(2009)
Article
Multidisciplinary Sciences
Michiko Miyazaki, Hideyuki Takahashi, Matthias Rolf, Hiroyuki Okada, Takashi Omori
SCIENTIFIC REPORTS
(2014)
Article
Computer Science, Artificial Intelligence
Matthias Rolf, Jochen J. Steil
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2014)
Proceedings Paper
Computer Science, Artificial Intelligence
Matthias Rolf, Minoru Asada
5TH INTERNATIONAL CONFERENCE ON DEVELOPMENT AND LEARNING AND ON EPIGENETIC ROBOTICS (ICDL-EPIROB)
(2015)
Proceedings Paper
Computer Science, Artificial Intelligence
Matthias Rolf
2013 IEEE THIRD JOINT INTERNATIONAL CONFERENCE ON DEVELOPMENT AND LEARNING AND EPIGENETIC ROBOTICS (ICDL)
(2013)
Proceedings Paper
Automation & Control Systems
Matthias Rolf, Jochen J. Steil
2012 IEEE/RSJ INTERNATIONAL CONFERENCE ON INTELLIGENT ROBOTS AND SYSTEMS (IROS)
(2012)