- Robot Manipulation and Learning
- Muscle activation and electromyography studies
- Robotics and Automated Systems
- Robotic Path Planning Algorithms
- Hand Gesture Recognition Systems
- Robotics and Sensor-Based Localization
- Gaze Tracking and Assistive Technology
- Soft Robotics and Applications
- Robotic Locomotion and Control
- Prosthetics and Rehabilitation Robotics
- Teleoperation and Haptic Systems
- Motor Control and Adaptation
- Social Robot Interaction and HRI
- Modular Robots and Swarm Intelligence
- Tactile and Sensory Interactions
- Robotic Mechanisms and Dynamics
- Stroke Rehabilitation and Recovery
- Human Pose and Action Recognition
- Reinforcement Learning in Robotics
- Manufacturing Process and Optimization
- Advanced Vision and Imaging
- Human Motion and Animation
- Virtual Reality Applications and Impacts
- Advanced Manufacturing and Logistics Optimization
- Advanced Image and Video Retrieval Techniques
The University of Tokyo
2025
Nara Institute of Science and Technology
2015-2024
ORCID
2021
Malaysia University of Science and Technology
2018
Mitsubishi Electric (Japan)
2017
National Institute of Advanced Industrial Science and Technology
2013
National Archives and Records Administration
2004-2011
TechnipFMC (Germany)
2010
Teikyo University
2009
Institute of Science and Technology
2003-2007
This paper reports a method that uses humanoid robots as communication medium. There are many interactive under development, but due to their limited perception, interactivity is still far poorer than of humans. Our approach in this limit robots' purpose non-interactive medium and look for way attract people's interest the information convey. We propose using passive-social medium, which multiple converse with each other. conducted field experiment at train station eight days investigate effects
Healthy individuals modulate muscle activation patterns according to their intended movement and external environment. Persons with neurological disorders (e.g., stroke spinal cord injury), however, have problems in control due primarily inability pattern an appropriate manner. A functionality test at the level of individual muscles that investigates activity a interest on various motor tasks may enable muscle-level force grading. To date there is no extant work focuses application...
Bin picking is still a challenge in robotics, as patent recent robot competitions. These competitions are an excellent platform for technology comparisons since some participants may use state-of-the-art technologies, while others conventional ones. Nevertheless, even though points awarded or subtracted based on the performance frame of competition rules, final score does not directly reflect suitability technology. Therefore, it difficult to understand which technologies and their...
This paper proposes a novel method for hand-pose estimation that can be used vision-based human interfaces. The aim of this is to estimate all joint angles. In method, the hand regions are extracted from multiple images obtained by multiviewpoint camera system. By integrating these silhouette images, pose reconstructed as "voxel model." Then, angles estimated using three-dimensional model fitting between and voxel model. following two experiments were performed: (1) an simulator (2) real...
This paper introduces a multifingered robotic hand eNAIST-Handf and grip force control by slip margin feedback. The developed prototype finger of the NAIST-hand has new mechanism which all 3 motors can be placed inside palm without using wire-driven mechanisms. A method is proposed an incipient estimation. tactile sensor designed to active NAIST-Hand. consists transparent semispherical gel, embedded small camera, in order implement direct structure principle sensing are described.
Humanoid robots are expected to work in the human environment due their similarity shape, and they required achieve localization navigation autonomously. In this paper, indoor is realized based on a view-based approach using camera mounted head. case of humanoid robot, image blur swing walking crucial issue for matching during localization. The quantitative effect firstly investigated by utilizing motion capture system. Then method generate stable view sequence proposed detection optical...
Unlike conventional portable eye-tracking methods that estimate the position of mounted camera using 2-D image coordinates, techniques are proposed here present richer information about person's gaze when moving over a wide area. They also include visualizing scanpaths user with head-mounted device makes natural head movements. We employ Visual SLAM technique to pose and extract environmental information. When moves, method obtains 3-D point-of-regard. Furthermore, can be appropriately...
The demand for rehabilitation robots is increasing the upcoming aging society. Power-assisting devices are considered promising enhancing mobility of elderly and disabled people. Other potential applications muscle sports training. main focus this paper to control load selected muscles by using a power-assisting device, thus enabling ldquopinpointedrdquo motion support, rehabilitation, training explicitly specifying target muscles. By taking into account physical interaction between human...
Myoelectric prosthetic hands provide an appearance with five fingers and a grasping function to forearm amputees. However, they have problems in weight, appearance, cost. This paper reports on the Rehand, realistic electric hand created 3D printer. It provides that is same as cosmetic function. A simple link mechanism one linear actuator for printed parts achieve low cost, light ease of maintenance. An operating system based distance sensor natural operability equivalent myoelectric control...
Insertion of a flexible wire into hole is difficult task because the has tendency to buckle under external forces. Moreover, plastic deformation occurs on when force acting big. In this paper, we propose method inserting whereby acts and shape observed by stereo vision. A strategy transform deformed straight one insert presented. Finally, some experimental results are shown.
In this paper, a grip-force control of an elastic object is proposed based on visual slip margin feedback. When pressed and slid slightly rigid plate, partial slip, called "incipient slip" occurs the contact surface. The between plate estimated analytic solution Hertzian model. A 1 DOF gripper consists camera force sensor developed. can be from tangential measured by sensor, deformation radius area both camera. method, friction coefficient not explicitly needed. grip controlled direct...
Natural body gesturing and speech dialogue, is crucial for human-robot interaction (HRI) symbiosis. Real not only with one-to-one communication but also among multiple people. We have therefore developed a system that can adjust gestures facial expressions based on speaker's location or situation multi-party communication. By extending our already real-time gesture planning method, we propose adjustment suitable human demand through motion parameterization gaze planning, which allows...
As the retail industry keeps expanding and shortage of workers increasing, there is a need for autonomous manipulation products to support operations. The increasing amount customers in establishments such as convenience stores requires automation restocking, disposing straightening products. needs be time-efficient, avoid damaging beautify display In this paper, we propose robotic system restock shelves, dispose expired products, straighten environments. proposed mobile manipulator features...
To build smart human interfaces, it is necessary for a system to know user's intention and point of attention. Since the motion person's head pose gaze direction are closely related his/her attention, detection such information can be utilized natural intuitive interfaces. In this paper, we describe behavior recognition based on real-time stereo face tracking measure simultaneously. The key aspect our use vision together with simple algorithm which suitable processing. Our significantly...
Measuring the force on a flexible wire is difficult by using force/torque sensors when flexural rigidity of small because buckles and smaller than that can be measured sensor. This paper presents method calculating from its shape observed stereo vision. We propose inserting into hole in wall estimated force. Three experiments have been successfully carried out.
We explore how to represent, plan and learn robot pouring. This is a case study of complex task that has many variations involves manipulating non-rigid materials such as liquids granular substances. Variations pouring we consider are the type (such into glass or spreading sauce on an object), material, container shapes, initial poses containers target amounts. The learns select appropriate behaviors from library skills, tipping, shaking tapping, pour range variety containers. also...
Automated factories use deep-learning-based vision systems to accurately detect various products. However, training such requires manual annotation of a significant amount data optimize the large number parameters deep convolutional neural networks. Such is very time-consuming and laborious. To reduce this burden, we propose fully automated approach without any intervention. do this, associate one visual marker with object capture them in same image. if an image showing used for training,...
Summary Motor exploration, a key process in reinforcement-based motor learning, is triggered by suboptimal performance and leads to increased movement variability. Consequently, failure one trial may increase the likelihood of subsequent trials. If this pattern emerges, successive failures (SFs) are not merely events be avoided but part trial-and-error identifying optimal movements, which can facilitate learning. This study investigated whether how SFs occur above chance levels during...
This paper presents a MR compatible force sensor. Recently, MRI are widely used in various fields, from medical purpose to brain science, and simultaneous measurement of information enables more precise investigations obtained images living tissue. However, conventional metal sensors which contaminate the true signals cannot be environments. Furthermore, present have limitation accuracy, dynamic range multi axis sensibility. In this paper, an optical 2-axis sensor without any electronic...
This paper presents the design and development of robotic tweezers with a force- displacement-sensing capability driven by piezoelectric stack actuators. In order to satisfy sufficient stroke tip force for future medical operations, rhombus strain amplification mechanism is adopted. One serially connected actuators nested in end effector used as sensor. The force-displacement characteristics at outermost layer respect forces innermost PZT (i.e., forward model) obtained from lumped parameter...
The universal gripper has attracted attention due to its simple structure and advanced grasping ability for irregularly shaped objects. In this research, we propose a novel design granular-jamming-based which uses transparent filling semi-transparent membrane allow optical sensing detect both deformation of the object being grasped. By adjusting refractive index an oil mixture granular bodies, produced fully that allows use camera inside gripper. paper, present materials development our...