Shuoyu Wang

Adaptive controller for omni-directional walker: Improvement of dynamic model

An omni-directional walker (ODW) is being developed for both walking rehabilitation and walking support for people with walking disabilities. The ODW cannot accurately follow a training path planned by a physical therapist for walking rehabilitation due to the center-of-gravity shifts and load changes caused by users. To address this issue, a new center-dynamic model of the ODW is derived considering center-of-gravity shifts and load changes. An adaptive control method is shown. Comparison with a dual-loop proportional-integral (PI) controller in simulations shows that the proposed method improves the path tracking accuracy.

Three-dimensional motion analysis for gesture recognition using singular value decomposition

A gesture is a form of non-verbal communication in which visible bodily actions communicate particular messages, either in place of speech or together and in parallel with spoken words. Gestures are important in the communication between human and human. It will make a robot more human-friendly to enable it to communicate with human by gestures. Our research addresses to develop a method to recognize human gestures for a guide robot that can be used in hospitals, welfare facilities, and etc. In this paper, firstly, a novel 3D motion analysis algorithm for gesture recognition using singular value decomposition (SVD) is proposed. An experiment, in which five gestures is included, is carried out to testify the effectiveness of the algorithm. The experiment results indicate that the proposed algorithm is applicable for the guide robot to recognize human gestures in guidance.

Directional control of an omnidirectional walking support walker: adaptation to individual differences with fuzzy learning

We are developing a method to recognize a user’s directional intention to control an omnidirectional walker (ODW) according to the force interaction between the ODW and the user. Since the characteristics in the force interaction are different among persons especially for those with walking difficulty, a fuzzy learning method is developed in this study to adapt to the individual difference in forearm pressures in order to improve the usability of the method. The experiment results show that fuzzy learning can significantly improve the accuracy of recognition by updating the fuzzy rules according to the characteristics in the force interaction. Graphical Abstract

Knowledge Acquisition Method Based on Singular Value Decomposition for Human Motion Analysis

The knowledge remembered by the human body and reflected by the dexterity of body motion is called embodied knowledge. In this paper, we propose a new method using singular value decomposition for extracting embodied knowledge from the time-series data of the motion. We compose a matrix from the time-series data and use the left singular vectors of the matrix as the patterns of the motion and the singular values as a scalar, by which each corresponding left singular vector affects the matrix. Two experiments were conducted to validate the method. One is a gesture recognition experiment in which we categorize gesture motions by two kinds of models with indexes of similarity and estimation that use left singular vectors. The proposed method obtained a higher correct categorization ratio than principal component analysis (PCA) and correlation efficiency (CE). The other is an ambulation evaluation experiment in which we distinguished the levels of walking disability. The first singular values derived from the walking acceleration were suggested to be a reliable criterion to evaluate walking disability. Finally we discuss the characteristic and significance of the embodied knowledge extraction using the singular value decomposition proposed in this paper.

A novel direction control method for walking support with an omnidirectional walker

Safe and convenient walking support machines are strongly desired for the people suffering from walking disabilities. We have developed an omnidirectional walker (ODW) for walking support. In walking support, it is necessary to control the ODW following the users directional intention. In this paper, a novel interface is proposed to recognise the users directional intention according to the forearm pressures. The forearm pressures exerted to the ODW by the user with wrists and elbows are measured by four force sensors embedded in the ODWs armrest. The relationship between the forearm pressure and the directional intention was extracted as fuzzy rules and an algorithm was proposed for directional intention identification based on distance–type fuzzy reasoning method. We conducted a path tracking experiment with the proposed method. The results show that the algorithm is applicable to control the ODW in walking support.

Acquisition of Embodied Knowledge on Gesture Motion by Singular Value Decomposition

Communication is classified in terms of verbal and nonverbal information. We discuss an acquisition method of knowledge from nonverbal information. In particular, a gesture is an efficient form of nonverbal communication as well as in verbal ways, and we formulate here a method that measures similarity and estimation between gestures. A gesture includes human embodied knowledge, and therefore the visible bodily actions can communicate particular messages. However, we have infinite patterns for gesture, determined by personality. Recently, the singular spectrum analysis method is utilized as an attractive method. In this paper, we propose a new method for acquiring embodied knowledge from time-series data on gestures using singular value decomposition. The motion behavior is categorized into several clusters with similarity and estimation between interval time-series data. We discuss the usefulness of the proposed method using an example of gesture motion.

A novel imitation approach on human's obstacle avoidance ability considering knowledge radius

In order to apply humans action intelligence to improving the path and trajectory planning of autonomous mobile robots, we focus on the imitation of humans obstacle avoidance ability. Based on a developed teleoperation system, humans adaptive operation data considering local and global information on complex environment can be obtained. In this paper, from the perspective of emphasizing knowledge usage, we propose a novel imitation approach on humans obstacle avoidance ability by using the distance-type fuzzy reasoning method with knowledge radius. Firstly, the fuzzy rules representing operators obstacle avoidance ability are effectively obtained from operation data by means of the learning algorithm for distance-type fuzzy reasoning method; then the imitation of humans obstacle avoidance is achieved by combining effective knowledge radius in the reasoning process. Moreover, the properties of knowledge radius are summarized and will provide further guideline to the search of knowledge radius for specific problem. Finally, the effectiveness of proposed method is illuminated through any trial experiment

Directional control of an omnidirectional walker for walking support with forearm pressures

Walking is a fundamental human ability necessary for everyday life. We have developed an omnidirectional walker (ODW) for walking support to those who have walking disabilities. It is necessary for the ODW to know which direction the user is intending to go during walking support. A novel interface is proposed for the ODW to recognize directional intention according to the users forearm pressures which are measured by force sensors embedded in the armrest. The relationship between forearm pressures and directional intention was extracted as fuzzy rules and an algorithm is proposed for directional intention identification based on distance type fuzzy reasoning method. In this paper, we conduct walking support experiments with the proposed method. The results show that the algorithm is applicable to directional control in walking support.

Speed dependency of cerebral blood volume changes during visual cognitive activation measured with NIRS

In this study, multichannel near-infrared spectroscopy (NIRS) was used to measure the change in cerebral hemoglobin concentrations during an incomplete-letter recognition task consisting of a fast section and a slow section. In the task, 85% of the black pixels in the letter images were erased. The display period of the erased letters in the fast section and in the slow section were set to 1000 ms and 3000 ms, respectively, during which the subjects were instructed to respond verbally. Fourteen male volunteers participated in the experiment. Activations in the occipital and bilateral frontal areas were simultaneously examined with three 24-channel probes. The results demonstrated that the oxygenated hemoglobin concentration (oxyHb) during the fast section was relatively higher than that during the slow section. Furthermore, the oxy-Hb difference between the two sections in the occipital areas was more significant than that in the bilateral frontal areas, indicating that for the recognition task in this study, speed change affects the activations in the occipital areas more than those in the frontal areas.

Measurement and Analysis of Brain Activation during a Driving Task

As we know, human driving behavior is strongly relevant to cognitive characteristics. If such characteristics could be clarified, it will contribute to exploration of those issues for cognitive process and even the development of modern traffic tools. This study will address to the driving characteristics by means of measurement and analysis of cognitive state inside brain. In this paper, a relatively new method of multi-channel near-infrared spectroscopy (NIRS) was used to investigate the brain activation by independently manipulating the cognitive demand in a driving simulator. One conclusion was drawn: in spite of changing traffic scene, there seems to be relatively stable effect in a number of collaborating areas of brain which each have multiple relative specializations and engage in extensive inter-area collaborations. Concretely, left brain plays an initiative pole, while right brain closely follows towards the activation degree of left brain. Consequently, there is a balanced tendency of symmetrical activation for left and right brain, suggesting that an easier model may be feasible for modeling driving cognitive process that operates on different levels of environment.