Extraction and implementation of muscle synergies in neuro-mechanical control of upper limb movement

This work faces the redundancy problem, a central concern in robotics, in a particular force-producing task by using muscle synergies to simplify the control. We extracted muscle synergies from human electromyograph signals and interpreted the physical meaning of the identified muscle synergies. Based on the human analysis results, we hypothesized a novel control framework that can explain the mechanism of the human motor control. The framework was tested in controlling a pneumatic-driven robotic arm to perform a reaching task. This control method, which uses only two synergies as manipulated variables for driving antagonistic pneumatic artificial muscles to generate desired movements, would be useful to deal with the redundancy problem; thus, suggesting a simple but efficient control for human-like robots to work safely and compliantly with humans.     

Face relighting using discriminative 2D spherical spaces for face recognition

As part of the face recognition task in a robust security system, we propose a novel approach for the illumination recovery of faces with cast shadows and specularities. Given a single 2D face image, we relight the face object by extracting the nine spherical harmonic bases and the face spherical illumination coefficients by using the face spherical spaces properties. First, an illumination training database is generated by computing the properties of the spherical spaces out of face albedo and normal values estimated from 2D training images. The training database is then discriminately divided into two directions in terms of the illumination quality and light direction of each image. Based on the generated multi-level illumination discriminative training space, we analyze the target face pixels and compare them with the appropriate training subspace using pre-generated tiles. When designing the framework, practical real-time processing speed and small image size were considered. In contrast to other approaches, our technique requires neither 3D face models nor restricted illumination conditions for the training process. Furthermore, the proposed approach uses one single face image to estimate the face albedo and face spherical spaces. In this work, we also provide the results of a series of experiments performed on publicly available databases to show the significant improvements in the face recognition rates.     

Straight-through microchannel devices for generating monodisperse emulsion droplets several microns in size

The authors recently proposed a promising technique for producing monodisperse emulsions using a straight-through microchannel (MC) device composed of an array of microfabricated oblong holes. This research developed new straight-through MC devices with tens of thousands of oblong channels of several microns in size on a silicon-on-insulator plate, and investigated the emulsification characteristics using the microfabricated straight-through MC devices. Monodisperse oil-in-water (O/W) and W/O emulsions with average droplet diameters of 4.4–9.8 μm and coefficients of variation of less than 6% were stably produced using surface-treated straight-through MC devices that included uniformly sized oblong channels with equivalent diameters of 1.7–5.4 μm. The droplet size of the resultant emulsions depended greatly on the size of the preceding oblong channels. The emulsification process using the straight-through MC devices developed in this research had very high apparent energy efficiencies of 47–60%, defined as (actual energy input applied to droplet generation/theoretical minimum energy input necessary for making droplets) × 100. Straight-through MC devices with numerous oblong microfluidic channels also have great potential for increasing the productivity of monodisperse fine emulsions.     

FDCNet: filtering deep convolutional network for marine organism classification

Convolutional networks are currently the most popular computer vision methods for a wide variety of applications in multimedia research fields. Most recent methods have focused on solving problems with natural images and usually use a training database, such as Imagenet or Openimage, to detect the characteristics of the objects. However, in practical applications, training samples are difficult to acquire. In this study, we develop a powerful approach that can accurately learn marine organisms. The proposed filtering deep convolutional network (FDCNet) classifies deep-sea objects better than state-of-the-art classification methods, such as AlexNet, GoogLeNet, ResNet50, and ResNet101. The classification accuracy of the proposed FDCNet method is 1.8%, 2.9%, 2.0%, and 1.0% better than AlexNet, GooLeNet, ResNet50, and ResNet101, respectively. In addition, we have built the first marine organism database, Kyutech10K, with seven categories (i.e., shrimp, squid, crab, shark, sea urchin, manganese, and sand).     

Spatial indexing of high-dimensional data based on relative approximation

We propose a novel index structure, the A-tree (approximation tree), for similarity searches in high-dimensional data. The basic idea of the A-tree is the introduction of virtual bounding rectangles (VBRs) which contain and approximate MBRs or data objects. VBRs can be represented quite compactly and thus affect the tree configuration both quantitatively and qualitatively. First, since tree nodes can contain a large number of VBR entries, fanout becomes large, which increases search speed. More importantly, we have a free hand in arranging MBRs and VBRs in the tree nodes. Each A-tree node contains an MBR and its children VBRs. Therefore, by fetching an A-tree node, we can obtain information on the exact position of a parent MBR and the approximate position of its children. We have performed experiments using both synthetic and real data sets. For the real data sets, the A-tree outperforms the SR-tree and the VA-file in all dimensionalities up to 64 dimensions, which is the highest dimension in our experiments. Additionally, we propose a cost model for the A-tree. We verify the validity of the cost model for synthetic and real data sets. Edited by T. Sellis. Received: December 8, 2000 / Accepted: March 20, 2002 Published online: September 25, 2002     

A stiffness adjustment mechanism maximally utilizing elastic energy of a linear spring for a robot joint

This paper proposes a mechanism that adjusts mechanical stiffness around a robot joint and utilizes whole elastic energy of an elastic element. The proposed mechanism consists of a lead screw mechanism, a linear spring, and wires. The lead screw mechanism moves a nut of the lead screw mechanism to change a bending point of the wire, which connects the linear spring and the lead screw mechanism. Then, moment arm and ratio of joint rotation to extension of the spring are varied. As a result, joint stiffness is adjusted. Because this mechanism does not apply tension to the spring for the stiffness adjustment, whole elastic energy of the spring can be utilized for joint rotation. This utilization can minimize weight and size of the elastic element. Additional advantages of the proposed mechanism are mechanical simplicity, wide range of adjustable stiffness, and no energy consumption for keeping constant stiffness. We analyze characteristics of the proposed mechanism and compare with other mechanisms in detail. Device development and experimental results are provided for demonstrating the effectiveness of the proposed mechanism.     

Solving fuzzy shortest path problems by neural networks

In this paper, we introduce the neural networks for solving fuzzy shortest path problems. The penalization of the neural networks is realized after transforming into crisp shortest path model. The procedure and efficiency of this approach are shown with numerical simulations.     

Interactive multiple objective linear programming system implemented on a microcomputer

Linear programming is one of the most widely used Operations Research/Management Science techniques. Recently, multiple objective decision making has been well established as a practical approach to seek a satisfactory solution to a decision making problem. Much attention has been focused on a microcomputer as an economical management tool.

In this paper we propose an interactive goal attainment method using the eigenvector algorithm for solving a multiple objective linear programming problem interactively on microcomputers. In the software package Micro-LPS based on the method proposed, we design a conversational and user-friendly system in which the user commands are involved.     

Component-based robot system design for grasping tasks

The paper presents a robot system design with highly reusable components for a component-based robot system for manipulation tasks. The robot system is designed based on the analysis of manipulation tasks using a unified modeling language use case diagram. For a service robot with locomotion and manipulation mechanisms, reusability of robot system components is improved by adopting the proposed design. Our structure consists of scenario, task, robot information management server, data analyzer, sensor hardware controller, skill, and motion hardware controller on a component-based robot system. Based on the proposed robot system, we implemented a component-based robot system and subsequently realized a grasping motion by a service robot.     

Reversible computing and cellular automata—A survey

Reversible computing is a paradigm where computing models are defined so that they reflect physical reversibility, one of the fundamental microscopic physical property of Nature. In this survey/tutorial paper, we discuss how computation can be carried out in a reversible system, how a universal reversible computer can be constructed by reversible logic elements, and how such logic elements are related to reversible physical phenomena. We shall see that, in reversible systems, computation can often be carried out in a very different manner from conventional (i.e., irreversible) computing systems, and even very simple reversible systems or logic elements have computation- or logical-universality. We discuss these problems based on reversible logic elements/circuits, reversible Turing machines, reversible cellular automata, and some other related models of reversible computing.