The 12th COE Postdoctoral and Doctoral Researchers
Technical Presentation

Program (20 mins each: 15 mins presentation and 5 mins discussion)

1. "AR-Based Assistance System for Search of Disaster Victims Using Teleoperated Unmanned Helicopter"
   Masanao KOEDA (Robotics Laboratory : D3)
   小枝 正直 (ロボティスクス講座 : D3)
   

   [Abstract]
   In this presentation, we introduce an AR-based assistance system for search operations of disaster victims using an unmanned helicopter which is mounted an omnidirectional camera, a GPS, and a gyroscope. Earthquakes and other natural disasters which strike in urban areas make great damage, and swift actions to rescue disaster victims is needed. This system assists the search operation by showing annotations consist of virtual objects and string which are overlaid on a perspective image generated from an omnidirectional image. Using the developed system, we conducted experiments on Heijyo palace site to search persons around the helicopter. When the helicopter was flying at the attitude of 6[m], human figure 30[m] distant was found from acquired images. This experimental result demonstrated the feasibility of developed system.
   
   Top

2. "Driver's degree of dependence on collision avoidance systems"
   黒崎 章 (システム制御・管理講座 : D1)
   Akira KUROSAKI (Systems and Control Laboratory : D1)
   

   [Abstract]
   In recent years, due to the development of information-communication technology, much progress has been made in the ITS used in cars. Recent years have seen an increased number of cars on the market featuring the electronic toll collection and adaptive cruise control functions. Therefore, there is a need to improve the safety and reliability of these systems and to reduce the driver’s workload while driving. However, if the driver is too dependent on collision avoidance assist system, there is a possibility that the driver could not avoid an accident without a system, if it is even under normal circumstances. In this research, we perform an experiment on a collision avoidance assist system using a driving simulator. Based on the analysis of the driver’s behavior and conduct, we evaluate the driver’s degree of dependence on the system.
   We also investigate the correlation between the driver’s degree of dependence and the driver’s conduct and behavior.
   
   Top

3. "3D Modeling of Outdoor Environments by Integrating Omnidirectional Range and Color Images "
   Toshihiro ASAI (Vision and Media Computing Lab. : D1)
   浅井 俊弘 (視覚情報メディア講座 : D1)
   

   [Abstract]
   We propose a 3D modeling method for wide area outdoor environments which is based on integrating omnidirectional range and color images. In the proposed method, outdoor scenes can be efficiently digitized by an omnidirectional laser rangefinder and an omnidirectional multi-camera system (OMS). In order to register multiple range images stably, the points on the plane portions detected from the range data are used in registration process. The 3D model which is obtained by registration of range data is mapped by the texture selected from omnidirectional images in consideration of the resolution of the texture and occlusions of the model. In experiments, we have carried out 3D modeling of our campus with the proposed method.
   
   
   Top

4. "Collecting Evaluative Expressions for Opinion Extraction"
   Nozomi KOBAYASHI (Computational Linguistics Laboratory : D1)
   小林 のぞみ (自然言語処理学講座 : D1)
   

   [Abstract]
   Automatic extraction of human opinions from Web documents has been receiving increasing interest. To automate the process of opinion extraction, having a collection of evaluative expressions such as the seats are comfortable would be useful. However, it can be costly to manually create an exhaustive list of such expressions for many domains, because they tend to be domain-dependent. Motivated by this, we have been exploring ways to accelerate the process of collecting evaluative expressions by applying a text mining technique. This poster proposes a semi-automatic method that uses particular cooccurrence patterns of evaluated subjects, focused attributes and value expressions.
   
   
   Top

5. "Information Extraction and Sentence Classification applied on Clinical Trial MEDLINE Abstracts"
   Kazuo HARA (Computational Linguistics Laboratory : D1)
   原 一夫 (自然言語処理学講座 : D1)
   

   [Abstract]
   We report experimental results on applying information extraction (IE) methodology to extract clinical trial design information of “Compared Treatment”, “Endpoint”and “Patient Population” from clinical trial MEDLINE abstracts, for the requirements from the practice of Evidence-Based Medicine (EBM). From these results, we have come to see this problem as one that can be decomposed into a sentence classification subtask and an IE subtask. By classifying sentences from clinical trial abstracts and only performing IE on sentences that are most likely to contain relevant information, we hypothesize that the accuracy of information extracted from the abstracts can be increased. In order to test this theory, we performed an experiment applying state-of-the-art sentence classification technique on the clinical trial abstracts and evaluated its potential in the original task of the extraction of clinical trial design information.
   
   
   Top

6. "Kernel based Link analysis"
   Takahiko ITO (Computational Linguistics Laboratory : D1)
   伊藤 敬彦 (自然言語処理学講座 : D1)
   

   [Abstract]
   The application of kernel methods to link analysis is explored. We argue that a family of kernels on graphs provides a unified perspective on the three measures proposed for link analysis: relatedness and global importance. The framework provided by the kernels establishes relative importance as an intermediate between relatedness and global importance, in which the bias between relatedness and importance is naturally controlled by a parameter characterizing individual kernels.
   
   
   Top

NAIST Graduate School of Information Science