We propose a scheduling method called SCQ (Smoothly Changing Queue) which can control service rate by bulk size of video streaming services such as IPTV and VoD. Since SCQ allows queue length to change smoothly, video streaming services can be stably provided with low jitter. Queueing analysis results show that SCQ can more stably deliver video streaming with low jitter and loss than existing AQMs or queue length-based rate control methods.

In this paper, we address latency issue in Xen-ARM virtual machines. Despite the advantages of virtualization in mobile systems, the current Xen-ARM is difficult to apply to mobile devices because it has unpredictable I/O latency. This paper analyzes the latency of incoming packet handling in Xen-ARM, and presents how virtualization affects the latency in detail. To make the latency predictable, firstly, we modify Xen-ARM scheduler so that the driver domain can be promptly scheduled by the hypervisor. Secondly, we introduce additional paravirtualization of guest OS that minimizes non-preemptible code path. With our enhancements, 99% of incoming packets are predictably handled within one millisecond at the destined guest OS, which is a feasible time bound for most soft real-time applications.

Two fast quarter-pixel motion estimation algorithms using the integer motion vector (MV) already obtained by any fast motion estimation algorithm are proposed, which perform MV search only in half-pixel and quarter-pixel unit. The proposed algorithms search MV about 4-9 times as fast as the full search in quarter-pixel unit, while the PSNR is slightly degraded.

To improve the efficiency of spectrum utilization, cognitive radio systems attempt to use temporarily unoccupied spectrum which is referred to as a spectrum hole. To this end, QoS (Quality of Service) is one of the most important issues in practical cognitive radio systems. In this article, an efficient spectrum management scheme using self-reserving spectrum is proposed to support QoS for cognitive radio users. The self-reservation of a spectrum hole can minimize service dropping probability by using the statistical characteristics of spectrum bands while using optimum amount of resources. In addition, it realizes seamless service for users by eliminating spectrum entry procedure that includes spectrum sensing, spectrum request, and spectrum grant. Performance analysis and intensive system level simulations confirm the efficiency of the proposed algorithms.

A new method for blind watermarking based on quantization is proposed. The proposed scheme embeds a watermark on the lowest wavelet subband in order to be robust. Experimental results demonstrate the robustness of the algorithm against compression and other image processing attacks.

We introduce an extended EZW coder that uses flexible zerotree coding of wavelet coefficients. A flexible parent-child relationship is defined so as to exploit spatial dependencies within a subband as well as hierarchical dependencies among multi-scale subbands. The new relationship is based on a particular statistics that a large coefficient is more likely to have large coefficients in its neighborhood in terms of space and scale. In the flexible relationship, a parent coefficient in a subband relates to four child coefficients in the next finer subband in the same orientation. If each of the children is larger than a given threshold, the parent extends its parentship to the neighbors close to its conventional children. A probing bit is introduced to indicate whether a significant parent has significant children to be scanned. This enables us to avoid excessive scan of insignificant coefficients. Also, produced symbols are re-symbolized into simple variable-length binary codes to remove some redundancy according to a pre-defined rule. As a result, the wavelet coefficients can be described with a small number of binary symbols. This binary symbol stream gives a competitive performance without an additional entropy coding and thus a fast encoding/decoding is possible. Moreover, the binary symbols can be more compressed by an adaptive arithmetic coding. Our experimental results are given in both binary-coded mode and arithmetic-coded mode. Also, these results are compared with those of the EZW coder.

This paper presents a vision-based human interface system that enables a user to move a target object in a 3D CG world by moving his hand. The system can interpret hand motions both in a frame fixed in the world and a frame attached to the user. If the latter is chosen, the user can move the object forward by moving his hand forward even if he has changed his body position. In addition, the user does not have to keep in mind that his hand is in the camera field of view. The active camera system tracks the user to keep him in its field of view. Moreover, the system does not need any camera calibration. The key for the realization of the system with such features is vision algorithms based on the multiple view affine invariance theory. We demon-strate an experimental system as well as the vision algorithms. Human operation experiments show the usefulness of the system.

Signal detectors generally utilize nonlinear statistics of an original observation rather than the original observation as it is. The sign statistic, a typical example of the nonlinear statistics, is the sign information of an observation and the sign detector relies only on the sign statistic. Since either detector might be of a better performance depending on the situation, it is quite important to determine which is the best performer among the detectors, based on the given situational information about noise and signal strength. In this letter, a qualitative analysis is presented that the correlation coefficients between the statistics and original observation can be used to predict the asymptotic performance of a detector utilizing one of the statistics, relative to the other detectors.

In this paper, we aim to investigate the potential usefulness of machine learning in image quality assessment (IQA). Most previous studies have focused on designing effective image quality metrics (IQMs), and significant advances have been made in the development of IQMs over the last decade. Here, our goal is to improve prediction outcomes of “any” given image quality metric. We call this the “IQM's Outcome Improvement” problem, in order to distinguish the proposed approach from the existing IQA approaches. We propose a method that focuses on the underlying IQM and improves its prediction results by using machine learning techniques. Extensive experiments have been conducted on three different publicly available image databases. Particularly, through both 1) in-database and 2) cross-database validations, the generality and technological feasibility (in real-world applications) of our machine-learning-based algorithm have been evaluated. Our results demonstrate that the proposed framework improves prediction outcomes of various existing commonly used IQMs (e.g., MSE, PSNR, SSIM-based IQMs, etc.) in terms of not only prediction accuracy, but also prediction monotonicity.

A zerotree image-coding scheme is introduced that effectively exploits the inter-scale self-similarities found in the octave decomposition by a wavelet transform. A zerotree is useful for efficiently coding wavelet coefficients; its efficiency was proved by Shapiro's EZW. In the EZW coder, wavelet coefficients are symbolized, then entropy-coded for further compression. In this paper, we analyze the symbols produced by the EZW coder and discuss the entropy for a symbol. We modify the procedure used for symbol-stream generation to produce lower entropy. First, we modify the fixed relation between a parent and children used in the EZW coder to raise the probability that a significant parent has significant children. The modified relation is flexibly modified again based on the observation that a significant coefficient is more likely to have significant coefficients in its neighborhood. The three relations are compared in terms of the number of symbols they produce.

In this paper, we analyze wavelet-based coding in a rate-distortion (R-D) sense by using Laplacian and Markov models and verify the results with the performance of the typical embedded coders, EZW and SPIHT, and the non-embedded coder implemented here. Laplacian represents the probability density function (pdf) of wavelet coefficients and Markov statistical dependency within and among subbands. The models allow us to easily understand the behavior of a thresholding and quantization part and a lossless coding part and associate the embedded coders with the nonembedded coder, which is the point the paper approaches. The analytic results are shown to coincide well with the actual coding results.

In this paper, we propose a text corpus design method for a Korean stereo super-wideband speech database. Since a small-sized text corpus for speech coding is generally required for speech coding, the corpus should be designed to comply with the pronunciation behavior of natural conversation in order to ensure efficient speech quality tests. To this end, the proposed design method utilizes a similarity measure between the phoneme distribution occurring from natural conversation and that from the designed text corpus. In order to achieve this goal, we first collect and refine text data from textbooks and websites. Next, a corpus is designed from the refined text data based on the similarity measure to compare phoneme distributions. We then construct a Korean stereo super-wideband speech (K-SW) database using the designed text corpus, where the recording environment is set to meet the conditions defined by ITU-T. Finally, the subjective quality of the K-SW database is evaluated using an ITU-T super-wideband codec in order to demonstrate that the K-SW database is useful for developing and evaluating super-wideband codecs.

In this paper, we propose a bandwidth-scalable stereo audio coding method based on a layered structure. The proposed stereo coding method encodes super-wideband (SWB) stereo signals and is able to decode either wideband (WB) stereo signals or SWB stereo signals, depending on the network congestion. The performance of the proposed stereo coding method is then compared with that of a conventional stereo coding method that separately decodes WB or SWB stereo signals, in terms of subjective quality, algorithmic delay, and computational complexity. Experimental results show that when stereo audio signals sampled at a rate of 32 kHz are compressed to 64 kbit/s, the proposed method provides significantly better audio quality with a 64-sample shorter algorithmic delay, and comparable computational complexity.

We introduce an entropy coding method to enhance the compression efficiency of JPEG. Because run-length coding and early-termination work more effectively for longer zero sequences, we extract ones and negative ones from the coefficients and reduce the magnitude of all coefficients by one. The extracted coefficients are encoded with a designated entropy coding method. The proposed method can transmit images in two parts progressively, where the first contains JPEG-compatible image with a small amount of degradation and the second is used to add fine details. Our method improves the compression ratio by more than 5% without sacrificing the efficiency of JPEG.