A5021260328- Robotics and Sensor-Based Localization
- Advanced Vision and Imaging
- Robotic Path Planning Algorithms
- Optical measurement and interference techniques
- 3D Surveying and Cultural Heritage
- Advanced Image and Video Retrieval Techniques
- Image Enhancement Techniques
- Balance, Gait, and Falls Prevention
- Indoor and Outdoor Localization Technologies
- Autonomous Vehicle Technology and Safety
- Modular Robots and Swarm Intelligence
- Image Processing Techniques and Applications
- Infrastructure Maintenance and Monitoring
- Motor Control and Adaptation
- Underwater Vehicles and Communication Systems
- Robotics and Automated Systems
- Model Reduction and Neural Networks
- Video Surveillance and Tracking Methods
- Muscle activation and electromyography studies
- Stroke Rehabilitation and Recovery
- Underwater Acoustics Research
- Geophysical Methods and Applications
- Teleoperation and Haptic Systems
- Free Will and Agency
- Industrial Vision Systems and Defect Detection
The University of Tokyo
2016-2025
University of Miyazaki
2022-2025
National Institute of Fitness and Sports in Kanoya
2025
Yamagata Prefectural Central Hospital
2024
Massachusetts General Hospital
2024
Kanazawa University
2024
Chiba University
2023-2024
Kitasato University
2022-2023
Nara Institute of Science and Technology
2023
Bunkyo University
2022
Path planning field for autonomous mobile robot is an optimization problem that involves computing a collision-free path between initial location and goal location. In this paper, we present improved artificial potential based regression search (Improved APF-based RS) method which can obtain global sub-optimal/optimal efficiently without local minima oscillations in complete known environment information. We redefine functions to eliminate non-reachable problems, utilize virtual target...
We propose a new detection method to predict vehicle's trajectory and use it for detecting lane changes of surrounding vehicles. According the previous research, more than 90% car crashes are caused by human errors, main factor. Therefore, if change can be detected before vehicle crosses centerline, accident rates will decrease. Previously reported methods have problem frequent false alarms zigzag driving that result in user distrust safety support systems. Most cases abortion due presence...
Continuous deep learning architectures have recently re-emerged as Neural Ordinary Differential Equations (Neural ODEs). This infinite-depth approach theoretically bridges the gap between and dynamical systems, offering a novel perspective. However, deciphering inner working of these models is still an open challenge, most applications apply them generic black-box modules. In this work we "open box", further developing continuous-depth formulation with aim clarifying influence several design...
In this paper, we propose a motion-planning method of multiple mobile robots for cooperative transportation large object in three-dimensional environment. This task has various kinds problems, such as obstacle avoidance and stable manipulation. All these problems cannot be solved at once, since it would result dramatic increase the computational time. Accordingly, divided motion planner into global path local manipulation planner, designed them, integrated them. The aim was to integrate...
The sense of agency refers to the feeling that one is controlling events through one’s own behavior. This study examined how task performance and delay influence during continuous action accompanied by a goal. participants were instructed direct moving dot into square as quickly possible pressing left right keys on keyboard control direction in which traveled. interval between key press response (i.e., change) was manipulated vary difficulty. Moreover, assisted condition, computer ignored...
Sit-to-stand (STS) motion is an important daily activity, and many post-stroke patients have difficulty performing STS motion. Previous studies found that there are four muscle synergies (synchronized activations) in the of healthy adults. However, for patients, it unclear whether change which features primarily reflect motor impairment. Here, we use a machine learning method to demonstrate temporal two contribute hip rising balance maintenance impairment patients. Analyzing age-matched...
We introduce the framework of continuous--depth graph neural networks (GNNs). Graph ordinary differential equations (GDEs) are formalized as counterpart to GNNs where input-output relationship is determined by a continuum GNN layers, blending discrete topological structures and equations. The proposed shown be compatible with various static autoregressive models. Results prove general effectiveness GDEs: in settings they offer computational advantages incorporating numerical methods their...
In this paper, we propose a shape reconstruction and image restoration method for paper documents with curved surfaces or fold lines by using stereo vision system. Characters in images of thick book's pages acquired an scanner are difficult to recognize because they deformed under the influence surface. Therefore, 3-D surface is executed from result stereoscopic measurement putting book upward, flat recovered folded The validity proposed shown through experiments.
Colors of objects observed in underwater environments are different from those air. This is because the light intensity decreases with distance water by attenuation. Robots on ground or air usually recognize surrounding using images acquired cameras. The same is/will be true robots. However, recognition methods based image processing techniques may become invalid Therefore, we propose a color registration method images. proposed estimates where acquired, other words, parameters essential to...
This paper presents an effective improved artificial potential field-based regression search (improved APF-based RS) method that can obtain a better and shorter path efficiently without local minima oscillations in environment including known, partially known or unknown, static, dynamic environments. We redefine functions to eliminate problems, use wall-following methods for the robots escape non-reachable target problems. Meanwhile, we develop optimise planned path. The optimisation is...
Sense of agency (SoA) refers to the feeling controlling one's own actions, and experience external events with actions. The present study examined effect strength intentional effort on SoA. We manipulated using three types buttons that differed in amount force required depress them. used a self-attribution task as an explicit measure results indicate enhanced when action-effect congruency was unreliable. concluded importantly affects integration multiple cues affecting judgments causal...
The sense of agency refers to the feeling control over one's own actions, and through them, external events. This study examined effect modified visual feedback on body movements using virtual reality in healthy individuals whose motor was disturbed. Participants moved a object their right hand trace trajectory (Experiment 1) or leading target 2). Their disturbed by delay 1-kg weight attached wrist In offset conditions, presented at median point between desired position participants' actual...
In this paper, we propose a novel method to generate realistic acoustic datasets for forward-looking sonars. A sonar, which is also known as an camera, outperforms other imaging sensor when applied in underwater tasks it can provide more accurate and detailed information about the environment, even dark or turbid water. However, difficulty high cost of acquiring images real experiments encourage researchers consider generation simulated image datasets. particular, deep learning-based methods...
The Phase angle (PA) measured by bioelectrical impedance analysis (BIA) is a potential indicator of exercise performance. Owing to the lack studies on adolescent female athletes, this study aimed investigate relationship between whole-body and regional PA performance in basketball players. Forty-five players (16.6±0.6 years) participated study. Lean soft tissue (LST) PAs for (PAWB), upper limb (PAUL), lower (PALL) were assessed using BIA method. Participants performed maximal isometric knee...
In this paper, we propose a method for generating reaching motions the insertion of flat cables. Despite need to insert cables into sockets in circuit assembly various electronic devices, there has been little research on automating that are already fixed one side board. regard, focus generation posture grasping such Our uses deep reinforcement learning simulation environment, and features extracted from image pose manipulator used as states. For transfer environment real-world use...