Scale conversion telexistence is an extended concept of telexistence. It comprises avatars of different scales compared with the human scale. Immersive telexistence experience requires the avatar to realize the operator's behavior instantaneously as soon as possible. It is important to match the avatar's behavior with the operator's behavior because the difference between the operator and the avatar tends to hurt the immersiveness. However, the scale difference between an operator with the avatar makes behavior mismatch and lets the operator experience under different physical constant world. For instance, gravity changes under scale conversion telexistence. Therefore, our previous work reveals its importance termed equivalent gravity condition for walking using a small biped robot. Thus, in this demonstration, we use a small biped robot as the different-scale avatar allow us to experience as if we are within the small robot. Attendees experience several steps walking under different gravity conditions.

In this work, we proposed the control method that enable intuitive operation, “tsumori control”. Tsumori control is implemented using human discrete intention of motion included in continuous and intuitive motion output. The experience operating a robot using tsumori control will be demonstrated for our control system.

This paper presents an efficient object detection method from satellite imagery. Among a number of machine learning algorithms, we proposed a combination of two convolutional neural networks (CNN) aimed at high precision and high recall, respectively. We validated our models using golf courses as target objects. The proposed deep learning method demonstrated higher accuracy than previous object identification methods.

The Earth observation satellites have been monitoring the earth's surface for a long time, and the images taken by the satellites contain large amounts of valuable data. However, it is extremely hard work to manually analyze such huge data. Thus, a method of automatic object detection is needed for satellite images to facilitate efficient data analyses. This paper describes a new image feature extended from higher-order local autocorrelation to the object detection of multispectral satellite images. The feature has been extended to extract spectral inter-relationships in addition to spatial relationships to fully exploit multispectral information. The results of experiments with object detection tasks conducted to evaluate the effectiveness of the proposed feature extension indicate that the feature realized a higher performance compared to existing methods.

People carry around mobile devices all the time in their daily life and get various information from the Internet in various situations. When searching for information (content) by using mobile devices, users' activities (e.g., walking and standing) and their situations (e.g., commuting in the morning and going out downtown in the evening) often change and this change may affect their degree of concentration on the display of mobile devices and their information needs. Therefore, search systems should provide users with an amount of information suitable for their activities and with a type of information suitable for their situations. In this paper, we present the design and implementation of a content search system considering mobile users' activities and situations, which aims to reduce users' load of operations in content searching. Our system recognizes user's activities and switches between two kinds of content search systems according to the user's activity: the location-based content search system runs when the user is standing, while the menu-based content search system runs when the user is walking. Both systems present information based on the user's situation. We also introduce a user's activity recognition method for mobile devices. This method classifies user's activity into standing, walking, and running using the sensors equipped in mobile devices.

Smart-phone users can search for information about surrounding facilities or a route to their destination. However, it is difficult to get or search for information while walking because of low legibility. To address this problem, users have to stop walking or enlarge the screen. Our previously proposed system for smart-phone switches the information presentation policies in response to the user's context. In this paper we describe our context recognition mechanism for this system. This mechanism estimates user context from sensors embedded in a smart-phone. We use a Support Vector Machine for the context classification and compare four types of feature values consisting of FFT and 3 types of Wavelet Transforms. Experimental results show that recognition rates are 87.2 % with FFT, 90.9 % with Gabor Wavelet, 91.8 % with Haar Wavelet, and 92.1 % with MexicanHat Wavelet.

Passive running robot can run down shallow slope with no energy source without gravity. The robot can generate a stable limit cycle. Dynamics of passive running is very interesting target and important for developing the robot and understanding the human locomotion. By the way, running high jumper can jump very high. The point of jumping is to transfer horizontal velocity into lifting momentum through the hinged-movement. In this paper, we demonstrate the experimental passive running based on the hinged-movement. Moreover, we derive the fixed point of the passive running, and analyze its stability.