Hui-Huang Hsu

Asynchronous Migration of Process Replicas in a Cluster

Application processes have to be efficiently and reliably performed on servers in a cluster. A process is replicated to increase the reliability and availability. However, the more number of replicas of a process are performed, the more reliable and available the system is but the more amount of electric energy is consumed. In this paper, we take a process migration (MG) approach to energy-efficiently and reliably performing an application process on servers in a cluster. A process performed on a current server is migrated to another server if the server is expected to consume a smaller amount of electric energy to perform the process than the current server while the deadline constraint of the process is satisfied. In order to reliably perform a process, multiple replicas of the process are performed on different servers. We consider synchronous and asynchronous ways to migrate process replicas to servers. In the synchronous migration, every replica moves to another server on a same computation state. In the asynchronous migration, each replica by itself makes a decision on when the replica is migrated to another server.

An RFID-Based Reminder System for Smart Home

Some people tend to forget things when they leave home for work or school. It is desired that there would be a reminder system to automatically remind people what they might have forgotten to bring along just when they step outside their home. In this paper, we present a reminder system that uses the RFID technology to detect the objects that a user brings along. The system then provide a reminder object list to the user based on the history data collected from the same user and the events in the users calendar on that day. The list is to remind the user objects he/she might have forgotten at home. A feedback mechanism is also designed to lower the possibility of unnecessary reminding. The user can mark the objects that are not needed from the reminder object list such that these objects will not appear in the list again in the same situation. A prototype system is introduced in this paper.

A Strategic Decomposition for Adaptive Image Transmission

Progressive image transmission (PIT) transmits the most significant portion of a picture, followed by its less important portions. The mechanism can be used in Web- based applications while users are browsing images. However, most PIT methods use the same pixel interpolation scheme for the entire picture, without considering the differ- ences among image blocks. This paper analyzes the efficiency of pixel interpolation schemes and test several decomposition mechanisms. The contribution results in an adaptive image transmission scheme, which takes the differences of picture portions into consideration. Moreover, this study tested 200 pictures in different categories and pa- rameters. In consequence, the overall bit rates can be reduced significantly with good PSNR values and user satisfaction. The visual result is superior to progressive JPEG on both objective (quantitative) and subjective (human) measures. An error recovery pro- cedure is also implemented in case that the transmitted pictures need to be fully recov- ered.

Smartphone Indoor Localization with Accelerometer and Gyroscope

Many people get lost easily in public places if they are not familiar with the environment. It would be very helpful to have an indoor localization system. Many researches in indoor localization can be found in the literature. However, they usually need to deploy sensors in the environment. It is costly and inconvenient. In this research, we propose to develop a smartphone indoor positioning application based on accelerometer and gyroscope data. The PDR (Pedestrian Dead Reckoning) method is used to build this application. Calibration points are marked both on the floor ground and on the map of the application. The user first finds a calibration mark, stand on it and face the right direction. He/she then place the android icon (representing the user) on top of the calibration point. When he/she starts to move, the android icon also moves on the map following the real-time estimates of step length and orientation change for each step from accelerometer and gyroscope data, respectively. Preliminary results in walking distance and orientation estimation show high accuracy. The application seems promising and useful as long as a map and calibration marks are built in advance.

A cognition assessment authoring system for e-learning

With the rapidly development of distance learning and the XML (Extensible Markup Language) technology, metadata becomes an important item in an e-learning system. Today, many distance learning standards such as SCORM, AICC CMI, IEEE LTSC learning object meta-data (LOM), and IMS learning resource metadata XML binding specification, use metadata to tag learning materials, shareable content objects, and learning resources. However, most metadata is used to define learning materials and test problems. Few metadata is dedicated for assessment in learning. We propose an assessment metadata model for e-learning operations. With the support from the assessment metadata, we can collect information at the question cognition level, item difficulty index, item discrimination index, questionnaire style, and question style. The assessment analysis model provides individual questions, summary of test results, and analytical suggestions. The suggestions and results can tell teachers why a question is not suitable and how to correct it. Teachers can see the analysis of test result and fix problematic questions. With the cognition level analysis, teachers can avoid missing items in teaching. The mechanism developed also suggests an e-learning system, with adaptive learning content and individualized tests, as well as good advice for the teachers.

Real-time multiple tracking using a combined technique

This paper considers a real-time multi-object tracking algorithm for rigid and non-rigid objects. The major components of the tracking object system are extraction of the background image, adaptation of the background image, and identification of the extracted object. In each component, we improved existing methods without increasing the complexity of computation. The system was tested on our fish tank experiment that solves dynamic occlusions problems.

Adaptive image transmission by strategic decomposition

Progressive image transmission (PIT) transmits the most significant portion of a picture, followed by less important parts. The mechanism can be used in Web-based applications, while users are browsing images. Most PIT methods use the same pixel interpolation scheme for the entire picture, without considering the differences in different portions. We test several strategic decomposition mechanisms. Our contribution results in an adaptive image transmission mechanism, which takes the differences of pictures and picture portions into consideration. We tested 1500 pictures in different categories and parameters. The overall bit rates can be reduced significantly, with good PSNR values. An error recovery procedure is also implemented to guarantee the transmitted pictures that can be fully recovered.

An Interactive Video Course Platform with Authoring and Playback Systems

Interactive video was developed to enable different video experiences to users. By using the technology, video materials can be augmented to attract students. An interactive video course platform integrated with authoring and playback systems is proposed in this paper. Content providers can produce interactive video courses by using the interactive video authoring tool. Video files will be cut into several smaller segments called scenes as the basic unit in a video course. Interactive objects and multimedia resources can also be inserted to video courses to enhance variety. Users can play interactive video via the proposed playback platform. Interactions, messages and other multimedia objects will be displayed in scenes when users browse video courses. The integrated system was used in class room and received positive feedbacks. Interested readers are welcome to visit our demonstrations at http://member.mine.tku.edu.tw/www/ham/IVEL.wmv.

Motion Analysis for the Standing Long Jump

The standing long jump is a standard test for primary school students. It can be used to evaluate the development of basic sports skills of a child. This paper presents a system that can automatically detect the motion during a standing long jump from a video sequence. The silhouette of the jumper in the film is segmented from the background first for all frames. A stick model is applied to the silhouette found in the first frame. Then a GA-based search algorithm is used to find the stick models for the rest of the frames. The stick model points out the important joints of a person and can be used to represent the pose of the jumper in each frame. From the pose change in consecutive frames, we will be able to analyze the movement of the jumper.

People tracking in a multi-camera environment

This paper presents a video processing system that can track a human target across multi-cameras. The user can browse video clips from the system. When a target is identified by the user, the system can automatically track the target across different cameras. There are three main parts in the system. The first part is object segmentation by a Bayesian model. The second part is object tracking. The mean-shift algorithm is used here to track the interested target in the current camera. The third part is cross-camera tracking. When the target is diminishing or moves out of the shooting range of the current camera, the system looks for the target in the neighboring camera. This part is repeated until the target is out of the shooting range of the multi-camera system. When initializing the system, the user can set up the relative positions and shooting angles of the cameras. The developed system is suitable for analyzing video clips from multiple surveillance cameras to track possible crime suspects. Experimental results on three outdoor surveillance cameras are presented and the results show that our approach is useful.