|
3D-Odometry Using Tactile Wheels and Gyros: Localization Simulation of a Bike Robot |
|
Tomoyasu Ichimura(National Institute of Technology, Gunma College, Japan) |
| Odometry is a widely used method for estimating the location of a mobile robot on even terrain; the use of attitude sensors has expanded the application of this method to three-dimensional space. We, therefore, propose a novel 3D-odometry using tactile wheels and gyroscopes. In this study, we conducted simulations to investigate the accuracy of the proposed method in tracking the locality of a bike robot traveling on uneven terrain. |
|
|
Performance Estimation of Time-series Prediction on Environmental Factors Using Multiple Similar Time-series Due to Mean-field Analysis Via Random Couplings and Fields |
|
Yohei Saika(National Institute of Technology, Gunma College, Japan) |
| We construct a method of the time-series prediction via multiple time-series of the discrete variables which
are similar to a target on the basis of the Bayesian inference by using the maximizer of the posterior marginal (MPM)
estimate regarded as the multi-level full-connected spin system with random couplings and fields including correlation
coefficients between multiple time-series. Using the mean-field theory, we estimate the accuracy of the MPM estimate
at each time with respect to the set of the time-series generated by the artificial assumed true prior. Then, we find that
the optimal pe |
|
|
Evaluation of Transmission Quality of Visible Light Communication using Bit Error Rate Measurement |
|
Nobuo Sasaki, Hiroki Shimada, Syunpei Shimada, Hiroki Kobayashi(National Institute of Technology, Gunma College, Japan) |
| A bit error rate testing system has been developed and the transmission quality of visible light communication was studied. First, a bit error rate was measured without a disturbance light. A bit error rate clearly became worse when the disturbance light was inserted. The results showed that this testing system is possible to evaluate quantitatively the effect of disturbance light. Signal to noise ratio at the input of comparator was calculated, and measured bit error rate was plotted versus SNR. However, a difference between measurement and theoretical curve was observed. |
|
|
Statistical Analysis of Feminine Movements in Japanese Traditional Dance |
|
Nao Shikanai(Japan Women's University, Japan) |
| This study focused on Onna Odori of a male Japanese traditional dance master. Eight observers evaluated the movements. The results showed the male dancer’s movements were evaluated as being feminine. The movements were analyzed and compared to data from previous studies of skilled female dancers’ movements. The results showed significant differences in shoulder inclinations, the standard deviations of hip movements, and et al. This study tried to clarify the relevance of the observers’ impressions of the movements’ characteristics in relation to femininity using covariance structure analysis. |
|
|
Individual Identification by Gait Vibration Data Transmitted Floor |
|
Yuichi Nakamura, Kazuki Ito, Tatsuo Hasegawa, Shin Itami(National Institute of Technology, Anan College, Japan), Yohei Saika(National Institute of Technology, Gunma College, Japan), Masahiro Nakagawa(Nagaoka University of Technology, Japan) |
| The establishment of variety methods in the individual identification technology is desirable. In this study, it is considered that the method of individual identification from vibration data generated in gait. In the walking motion, the unique information such as the individual's body type and habit is contained. It is assumed that the unique information is reflected in the gait vibration data transmitted in the floor. The potential for application to individual identification is shown by extracting one step and analyzing the unique information from the gait vibration data. The Adaboost metho |
|
|
Probability Distribution of an Image Dictionary for Compressed Sensing |
|
Yuhei Ashida, Toshiaki Aida(Okayama University, Japan) |
| In this paper, we analytically derive the expression of the probability distribution followed by an image dictionary for compressed sensing, assuming that grey scale images are generated by the Gaussian model. This result enables us to directly generate a dictionary matrix for images with no edge, and can be the first step to analytical performance evaluation of image processing by compressed sensing. |
|