Tomohiro Umetani

Object recognition using appearance models accumulated into environment

Proposes a method of object recognition using appearance models accumulated into a RFID (radio frequency identification) tag attached to the environment. Robots recognize the object using appearance models accumulated in the tag on the object. If the robot fails in recognition, it acquires a model of the object and accumulates it to the tag. Since robots in the environment observe the object from different points of view at different time, various appearance models are accumulated as time passes. In order to accumulate many models, eigenspace analysis is applied. The eigenspace is reconstructed every time robots acquire the model. Experimental result of object recognition shows effectiveness of the proposed method.

Automated Handling of Construction Components Based on Parts and Packets Unification

ABSTRACT: This paper proposes a method of pose (position and orientation) fitting of construction components in a construction site for automated handling based on the relation between components (parts) and their information (packets). Robots can acquire the required information of the component via the environmentattached storages, such as RFID devices. When an ID reader identifies an ID device, it should take some pose in its communication range. This fact may bring the idea of estimating the pose of a component that carries the device. In this idea, only single device identification cannot fix the pose of the component. We define the conditions of the ID reader and ID devices for the pose fitting, and propose a fitting method with at least two different identifications where two devices are not attached to the same plane or parallel planes of the component. Several examples of pose fitting show the feasibility of our idea.

Parts and Packets Unification for Construction Automation and Robots

This paper proposes a method of configuration of database system for parts oriented construction. In parts oriented construction, robots acquire information related to the tasks from the server through ID attached to the construction components. This system aims (1) simplification in motion planning of the construction robots using information attached to components, (2) reconfiguration of data structure for process of construction tasks. The robot performs the tasks according to local planning based on the local information. The operation server reconfigure parts information based on task result by robots so that the next robot can obtain the required information from the composed components in the next phase. We call the parts integrated information “packets”, and call our idea “parts and packets unification.” The feasibility of our proposed method is also shown in the preliminary experiments where an RFID tag attached to a component is applied for data exchange.

Human-Robots Implicit Communication based on Dialogue between Robots using Automatic Generation of Funny Scenarios from Web

Numerous studies have examined communication robots that communicate with people, but it is difficult for robots to communicate with people smoothly. We call the communication style based on dialogue between robots as “human-robot implicit communication”. As described herein, we propose a Manzai-robots for which the interaction style is human-robot implicit communication based on an automatically generated scenario from web news. Our generated Manzai scenario consists of snappy patter and a misunderstanding of dialogue based on the four kinds of gap of structure of funny points. Our purpose is that people feel familiarity from smoothly human-robot communication using dialogue between robots based on a Manzai scenario. We conducted experiment of three kinds to assess (1) the effectiveness of automatic creation of Manzai scenario for the robots, (2) the effectiveness of the Manzai-robots as a media, and (3) the effectiveness of types of familiarity for Manzai-robots. Based on their results, we measured the familiarity and smooth communication of our Manzai-robots.

Detection of finger gesture using singular spectrum transformation

The purpose of this study is to detect finger movement using a singular spectrum transformation method. Human gesture recognition is essential for realizing natural user interfaces. However, constructing a robust, easily installable interface is extremely difficult. Our proposed method uses singular spectrum transformation to classify finger movements. This method robustly classifies gestures and behavior.

Development of Health Monitoring System based on Three-dimensional Imaging using Bio-signals and Motion Data

The purpose of this study is to develop a health monitoring system. This system represents human health condition and human motion simultaneously. This system obtains motion data by optical motion capture system. Electromyogram and electrocardiogram are used for health condition estimation. These are measured synchronously with motion data. Two experiments were performed to evaluate our system. In these experiments, muscle activity, motion of the heart and the autonomic nervous balance were represented as health conditions. This system contributes intuitive recognitions of health condition and high accurate estimation of health conditions.

Automatic generation of Japanese traditional funny scenario from web content based on web intelligence

Today there is much information and knowledge on the internet, and many studies have examined the extraction of many kinds of knowledge from the internet. In addition, numerous studies have examined entertainment robots that communicate with people, but it is difficult for robots to communicate smoothly with people. We specifically examine communication between robots based on dialogue. Here, we create a dialogue-based scenario for the robots to undertake automatically, but it is difficult because the dialogue requires knowledge of many kinds. We consider the use of the knowledge from the web and create scenarios automatically. As described herein, we propose a system that generates dialogue scenarios automatically from web news articles in real time. We used the Manzai metaphor, which is Japanese traditional humorous comedy in our system. Our generated Manzai scenario consists of snappy patter and a misunderstanding dialogue based on the gap of our structure of funny points. We create communication robots to amuse people with our generated humorous robot dialogue scenarios.

Pose Estimation of Construction Materials Using Multiple ID Devices for Parts and Packets Unification

: A method for characterizing the pose (position and orientation) of large objects using multiple identification (ID) devices, such as radio-frequency identification (RFID) devices, is proposed. Simultaneous management of construction materials and their data using ID devices for construction automation has been proposed. In this management scheme, workers or robots (1) acquire the ID of a device using an ID reader, (2) obtain construction material data using the ID, and (3) use the data to perform various tasks. In this scheme, the object pose is crucial because it can be changed by robots and workers who do not possess an ID reader or have not measured the pose previously. The proposed method uses the geometric relationship between the ID device attached to the object and the ID reader, and is based on the relationship between the relative displacements of multiple ID devices. The position and direction of each ID device and ID reader are measured with respect to the object and reference coordinate frames, respectively. Experimental results proved the feasibility of our proposed method.

Pose Estimation of Construction Materials Using Multiple ID Devices for Construction Automation

This paper describes a method of pose (position and orientation) of large object such as construction materials using multiple ID devices for construction automation. Object pose is crucial issue for construction automation, however, the pose can change by the robots and workers that do not have ID reader or measure the object pose. The proposed method supposes the position and direction of each ID device with respect to the object coordination frame and those of the ID reader with respect to the reference coordination frame. The paper focuses on the simplification of the estimation model to apply the method to the real environment. Experimental results show feasibility of our proposed method.

Identification of Partially-Exposed Metal Object

This paper describes a method of automated type and pose (position and orientation) identification of partially exposed metal objects that makes excavation of hazardous materials safe and efficient. The object pose is estimated by matching a model of the object to the area that is extracted from the range image using the characteristics of metal objects. Experimental results show feasibility of type and pose identification of partially exposed objects.