Changyeol Choi

Depth Scaling of Multiview Images for Automultiscopic 3D Monitors

In this paper, we present a depth scaling method for multiview images that could provide an effective control of stereoscopic depth range. Unlike the previous works that change a camera configuration, our proposed method utilizes depth data in order to carry out the scaling of a depth range requested by users. In particular, our method can deal with multiview images captured by multiple cameras, and can be expanded from stereoscopic to multiview images. Our experimental results tested on automultiscopic 3D displays show that the perceived depth is appropriately scaled according to a users preferred depth.

Load distribution strategies in cluster-based transcoding servers for mobile clients

The recent advance in wireless network technologies has enabled the streaming media service on the mobile devices such as PDAs and cellular phones. Since the wireless network has low bandwidth channels and mobile devices are actually composed of limited hardware specifications, the transcoding technology is needed to adapt streaming media to the given mobile devices. When large scale mobile clients demand the streaming service, load distribution strategies among transcoding servers highly impact on the total number of QoS streams. In this paper, the resource weighted load distribution strategy is proposed for the fair load balancing and the more scalable performance in cluster-based transcoding servers. Our proposed strategy is based on the weight of resources consumed for transcoding to classified client grades and the maximum number of QoS streams actually measured in transcoding servers. The proposed policy is implemented on cluster-based transcoding system. In experiments, we evaluate its fair load distribution and scalable performance according to the increase of transcoding servers.

Multiview Video Service Framework for 3D Mobile Devices

Multiview broadcasting has recently gained much attraction from academic and commercial fields because it can deliver the immersive viewing of natural scenes. Therefore, various researches implementing multi-view video broadcasting have been conducted with a variety of different system architectures. However, the previous systems have focused on wired broadcasting. On the contrary, our system is designed to be applied to mobile 3DTV broadcasting. We present the system adaptable to mobile clients where the load on the client is reduced as much as possible. MPEG-21 multiview DIA description schemes are presented for the efficient adaptation between the serer and client.

Thin-Client computing for supporting the qos of streaming media in mobile devices

Due to the limited resources in thin-clients on mobile devices and the large amount of computation for decoding streaming media, it is not easy to support the QoS of streaming media for thin-client on mobile environment. To solve the problems, the terminal servers would be charged for decoding the streaming media and thin-clients have a role to update only the changed areas in their screen. In this paper, an intelligent media player is proposed to provide the improved QoS for streaming media in thin-client computing. Since the proposed method reflects the intrinsic property of streaming media, it provides both the enhanced video quality and the audio streaming fully synchronized with image frames.

A Self-optimizing Mechanism of Location Aware Systems for Ubiquitous Computing

The mobility or highly interconnected and communicating devices and users has implications for the QoS in a ubiquitous computing environment. Therefore, it is important for location aware systems to detect location of mobile object correctly and Provide high quality services in ubiquitous environment. However, it is not easy that location aware systems offer highly reliable QoS to users because process strategies of location aware systems are limited by the capability according to the applied detection target objects. In this paper, we design an autonomic architecture which analyzes the location aware system condition and autonomously chooses the best appropriate process strategy. We also have simulated the Proposed architecture in order to verify its performance. The test results show us that the architecture using self-optimizing mechanism provides higher QoS to users in variable bandwidth.

Cloud Assisted P2P Live Video Streaming over DHT Overlay Network

Many works have attempted to solve the scalability, the availability, and the low-latency problems of peer-to-peer (P2P) live video streaming; yet, the problems still remain. While tree-based systems are vulnerable to churn, the mesh-based systems suffer from high delay and overhead. The DHT-aided chunk-driven overlay (DCO) [1] tried to tackle these problems by using the distributed hash table (DHT), which structures into a mesh-based overlay to efficiently share the video segment. However, DCO fully depends on the capacity of the users’ device which is small and unstable, i.e., the users’ device may leave and join the network anytime, and the video server’s bandwidth can be insufficient when the number of users joining the network suddenly increases. Therefore, cloud assist is introduced to overcome those problems. Cloud assist can be used to enhance the availability, the low-latency, and the scalability of the system. In this paper, the DHT is used to maintain the location of the streaming segments in a distributed manner, and the cloud server is used to assist other peers when the bandwidth which required for sharing the video segment is insufficient. The simulation results show that by using the threshold and cloud assist, the availability and the low-latency of the video segments, and the scalability of the network are greatly improved.

Adaptive Multi-view Video Service Framework for Mobile Environments

In this paper, we propose an adaptive multi-view video service framework suitable for mobile environments. The proposed framework generates intermediate views in near-realtime and overcomes the limitations of mobile services by adapting the multi-view video according to the processing capability of a mobile device as well as the user characteristics of a client. By implementing the most of adaptation processes at the server side, the load on a client can be reduced. H.264/AVC is adopted as a compression scheme. The framework could provide an interactive service with efficient video service to a mobile client. For this, we present a multi-view video DIA (Digital Item Adaptation) that adapts the multi-view video according to the MPEG-21 DIA multimedia framework. Experimental results show that our proposed system can support a frame rate of 13 fps for 320{\times}240 video and reduce the time of generating an intermediate view by 20 % compared with a conventional 3D projection method.

Multi-View video Processing : IVR, Graphics Composition, and Viewer

Multi-view video has recently gained much attraction from academic and commercial fields because it can deliver the immersive viewing of natural scenes. This paper presents multi-view video processing being composed of intermediate view reconstruction(IVR), graphics composition, and multi-view video viewer. First we generate virtual views between multi-view cameras using depth and texture images of the input videos. Then we mix graphic objects to the generated view images. The multi-view video viewer is developed to examine the reconstructed images and composite images. As well, it can provide users with some special effects of multi-view video. We present experimental results that validate our proposed method and show that graphic objects could become the inalienable part of the multi-view video.

An Efficient Checkpoint Protocol in Wireless Sensor Network for Reliability

The reliability concept of wireless sensor network is essential to get exactly actual data from the ubiquitous environment. A rollback technique for the self-healing helps to increase it. However, a fault can occur in wireless sensor network when to use a previous rollback technique because it is designed just for the local system. So, checkpoint protocols are suggested in order to use a rollback technique in the network without the fault. However, there is trade-off among performance overhead, power consumption, and memory overhead for each of protocols. Hence, we suggest a novel global checkpoint protocol, so called address log based protocol(ALBP), based on an asynchronous protocol. It is a platform based protocol to reduce power consumption, performance overhead, and memory overhead which are the most of consideration in wireless sensor network.

Design and implementation of a cache-conscious index manager for the Tachyon, a main memory DBMS

The main memory DBMS (MMDBMS) efficiently supports various database applications that require high performance since it employs main memory rather than disk as a primary storage. In this paper, we discuss the cache-conscious index manager of the Tachyon, a next generation MMDBMS. The index manager is an essential sub-component of a DBMS used to speed up the retrieval of objects from a large volume of a database in response to a certain search condition. Recently, the gap between the CPU processing and main memory access times is becoming much wider due to rapid advance of CPU technology. By devising data structures and algorithms that utilize the behavior of the cache in CPU, we are able to enhance the overall performance of MMDBMSs considerably. In this paper, we address the practical implementation issues and our solutions for them obtained in developing the cache-conscious index manager of the Tachyon. The main issues touched are (1) consideration of the cache behavior, (2) compact representation of an index entry, (3) support of variable-length keys, (4) support of multiple-attribute keys, (5) support of duplicated keys, and (6) definition of the system catalog for indexes. We also show the effectiveness of our approach through extensive experiments.