Future of Multimedia Systems and Applications

In Kameyama Lab., various research projects are going on under the same theme of Multimedia Information Distribution Systems. Below, you can find some of the on-going and completed projects.

Extraction and Description of Atmosphere Metadata using Bio-signals Today, various metadata are accompanied with contents. Metadata representing mood and genre of audiovisual contents are referred to as "Atmosphere metadata" in the international standards by TV-Anytime Forum. They are essentially important to realize personalization and recommendation systems for contents. In this project, we try to extract atmosphere metadata from user's emotion using bio-signals and to describe them for further usage in the system. For more detail	Automatic Image Annotation Project Automating the process of annotation of digital images is a crucial step towards efficient and effective management of this increasingly high volume of content. However, this is still a highly challenging task for the research community. One of the main bottle necks is the lack of integrity and diversity of features. We solve this problem by proposing to utilize 43 image features that cover the holistic content of the image from global to subject, background and scene. In our approach, saliency regions and background are separated without prior knowledge. Each of them together with the whole image are treated independently for feature extraction. Extensive experiments are designed to show the efficiency and effectiveness of our approach as well as giving detail analysis. For more detail
Contents Distribution Analysis Project As secondary information distribution (distribution from users to users) has been playing an important role in our society, We propose an information distribution model and equations which take account of human behavior in order to analyze the power of secondary information distribution more efficiently. Subsequently, the new proposed model simulates the phenomenon by applying the small-world network and its characteristics. The proposed model parameters and the small-world network structures are also analyzed to find their impacts on the effectiveness of the information distribution. We are also analyzing the information distribution on many real world networks by applying large network data set. For more detail	Khmer Language Processing Project As the amount of content and information in Khmer is rapidly growing, searching for relevant information is very difficult since there is no specific implementation on Khmer information retrieval (IR) and effective Khmer processing system yet. Therefore, this research aims to study on Khmer IR system based on query expansion (QE), word segmentation for Khmer Language. For more detail
Aquarium AR System Aquariums are very popular in the world. However, it's really frustrating when you try to find some fish and other creatures in the fish tank because they hide behind rocks and they have protective colors on them, usually. In this project, in order to solve it, we propose an AR system to find and track fish and creatures, and to display the corresponding detail information on smart phone. The most challenging task in the project is to find and recognize creatures in fish tank because they are non-rigid and their skin colors may change due to the light reflection in the fish tank. For more detail

ナビゲーション

• TopTOP PAGE
• ProfessorPROFESSOR
• ResearchRESEARCHES
• MemberMEMBERS
• ContactCONTACT US
• LectureLECTURE
• JapaneseENGLISH

copyright©2014, Kameyama Laboratory all rights reserved.