Bangkok University - Multimedia Intelligent Technology (BU-MIT is a research team led by Dr. Worawat Choensawat and Dr. Kingkarn Sookhanaphibarn, working at the School of Science and Technology. Our group's research interests are in the multimedia development and system support tools, and case studies of multimedia applications for education and humanities. Our current research topics are shown below:

 

Research Detail:

We have long been conducting research on the description and reproduction of body motion based on Labanotation, and developed a system called LabanEditor. Labanotation uses a symbolic description and it is said that the notation can even describe motions of the fine each finger of a dancer. However, since, in this case, the resulting score staff will become extremely complicated, only the basic description has usually been used. On the other hand, there are styles of motion particular to traditional dances, and if we restrict ourselves to handling these dances, the basic notation scheme is insufficient to describe these motions. Based on the above, we investigated a method of describing and reproducing CG animation of highly-stylized traditional dance using plain Labanotation.

http://mit.science.bu.ac.th/userfiles/image/thaidance.jpg

 

 

 

 

 

 

Research detail:

Nowadays, Motion Capture (MoCap) data has been widely used for producing animations in a variety of applications such as animated movies, video games, simulations, and in the research field of human body analysis. The production of 3D animated content requires very precise data because MoCap provides accurate and semantically rich data.

Similarity retrieval of MoCap data has received substantial and increasing attention for several years. For applications, such as film production, creators make an effort to select appropriate motions to generate high quality animation. Therefore, an efficient similarity retrieval system is very important for the effective reuse of MoCap data because the more similar motions matching a motion query, the more realistically animated characters can be produced.

 

 

 

 

 

Research detail:

This project focuses on the development of a computer-aided tool for automatically generating Labanotation scores from motion capture data named GenLaban. GenLaban can be implemented with low-cost equipment but an efficient method that allows users converting body motions to scores. The key components of GenLaban are the analysis of body motions, the quantization of body postures and the determination of body parts carrying the body weight. All the processes are under supervision of a Labanotation expert to ensure the notation meaning correctly as the use for the dance composition.

 

http://mit.science.bu.ac.th/userfiles/image/laban_mocap.jpg

 

 

 

 

 

Research detail:

This research presents a virtual reality (VR) system for fire evacuation training and a natural user interface via Kinect for imitating users’ movements, e.g. walking, running, step side way, turn and open/close a door, during the training. The incident is set in a realistic environment by imitating one campus building in Bangkok University, where there are robotic, computer, and research laboratories, lecture rooms, and staff offices. The fire starting can be randomly changed to let trainees learn the exit path through the building.

 

 

 

 

 

 

 

 

 

Research detail:

It is widely acceptable that Labanotation is a useful tool for human movement recording, choreography and dance training. Labanotation is one of the most common movement notation systems. The reason lies in the fact that the complexity of Labanotation and lack of experts, it is not an easy task to introduce this new learning method based on the notation scheme to the dance community in Thailand. To overcome the barrier, one of the solutions is to use a computer aided tool to help new learner in understanding the Labanotation. A game-based system is proposed by using Microsoft Kinect as a motion capturing of human movement device.  The more fun users have playing a learning game, the greater will be their actual cognitive learning gain.

 


 

1.     Detection and monitoring malicious applications on android mobile devices (funding by the national broadcasting and telecommunication commission : NBTC during 2016-2017)

a.     Static analysis for malware detection by using machine learning

b.     Behavioral detection of malicious code on mobile operating systems

2.     Virtual learning center of cyber-securities awareness for home-users

a.     Extracting information approach from on-line news

b.     Web-based retrieval systems for on-line news

c.      Web-based training systems for  cyber-securities awareness

d.     Game-based learning environment for children and youth

e.     Information Visualization of cyber-crimes

 

 

Research Detail:

1.     Detection and monitoring malicious applications on android mobile devices (funding by the national broadcasting and telecommunication commission : NBTC during 2016-2017)

 

a.     Static analysis for malware detection by using machine learning

This research focuses on a model for malware detection on mobile operating system based on machine learning technique. The objective is to reduce the risk of installing harmful application when the user did not update the anti-virus program in time. The proposed model is different to other anti-virus is that most of anti-virus software used virus signature to identify malware. However, the virus signature-based detection approach requires frequent updates of the virus signature dictionary. The signature-based approaches are not effective against new, unknown viruses while the proposed model based on machine learning can detect new malware even some parts of the code have been modified.

 

 

b.     Behavioral detection of malicious code on mobile operating systems

This research focuses on earlier approaches for dynamic analysis of application behavior as a means for detecting malware in the Android platform. The detector is embedded in an overall framework for collection of traces from an unlimited number of real users based on crowdsourcing. Our framework has been demonstrated by analyzing the data collected in the central server using two types of data sets: those from artificial malware created for test purposes, and those from real malware found in the wild.

 

 

 

 

2.     Virtual learning center of cyber-securities awareness for home-users

a.     Extracting information approach from on-line news

In many real-world scenarios, the ability to automatically classify documents into a fixed set of categories is highly desirable. Common scenarios include classifying a large amount of unclassified archival documents such as newspaper articles, legal records and academic papers. For example, newspaper articles can be classified as ‘crimes’, ‘cyber-crimes’,  ’features’, ’sports’ or ’news’. Other scenarios involve classifying of documents as they are created. Examples include classifying movie review articles into ’positive’ or ’negative’ reviews or classifying only blog entries using a fixed set of labels.

Natural language processing offers powerful techniques for automatically classifying documents. These techniques are predicated on the hypothesis that documents in different categories distinguish themselves by features of the natural language contained in each document. Salient features for document classification may include word structure, word frequency, and natural language structure in each document. This project looks specifically at the task of automatically classifying newspaper articles in “cyber-crimes” from news feeds.

 

mass-scale-crawls

 

b.     Web-based retrieval systems for on-line news

This project discusses a system for online new event detection in the domain of news articles on the web. This area is related to the Topic Detection and Tracking initiative. We evaluate two benchmark systems: The first like most current web retrieval systems, relies on term repetition to calculate document relatedness. The second attempts to perform conceptual indexing through the use of the WordNet thesaurus software. We propose a novel approach for the identification of breaking news stories, which uses a technique called lexical chaining. We believe that this technique will improve the overall performance of our web retrieval system by allowing us to encapsulate the context surrounding a word and disambiguate its senses.

 

c.      Web-based training systems for  cyber-securities awareness

Web-based training (sometimes called e-learning) is anywhere, any-time instruction delivered over the Internet to browser-equipped learners. There are two primary models of Web-based instruction: synchronous (instructor-facilitated) and asynchronous (self-directed, self-paced). Instruction can be delivered by a combination of static methods (learning portals, hyperlinked pages, screen cam tutorials, streaming audio/video, and live Web broadcasts) and interactive methods (threaded discussions, chats, and desk-top video conferencing).

 

d.     Game-based learning environment for children and youth

A new interest in the use of games for learning has emerged, and a number of claims are made with respect to the effectiveness of games in education. These educational games are considered as new instructional technology with great potential. The suggested positive outcomes and effects have been mentioned repeatedly. In this project, the educational game for teaching how to use Internet and playing social network are studied and developed in order to gain more insights into the conditions under which a game may be effective for learning. A systematic literature search in three databases was conducted. Some studies reported a positive effect on learning and motivation, but this is moderated by different learner variables and depends on different context variables. Next to this, the effectivity research on game-based learning is highly susceptible to a muddle of approaches, methodologies, and descriptions of gaming for educational purposes.

 

e.     Information Visualization of cyber-crimes

Information visualization or information visualization is the study of (interactive) visual representations of abstract data to reinforce human cognition. The abstract data include both numerical and non-numerical data, such as text and geographic information. The field of information visualization has emerged "from research in human-computer interaction, computer science, graphics, visual design, psychology, and business methods. This project focuses on displaying the statistics of cyber-crimes in a way of interesting presentation to audiences.