MPEG-4 is a standardization effort under the auspices of ISO being developed by MPEG (Moving Picture Experts Group), the committee that also developed the Emmy Award-winning standards known as MPEG-1 and MPEG-2. MPEG-4 addresses the "generic coding of audio-visual objects." In contrast to all other existing audio or video representation standards, MPEG-4 adopts an object-based approach for content description: the content is assumed to be constructed out of individual and independent entities called objects, which are coded and transmitted separately. These objects include, for example, rectangular or arbitrarily shaped natural video, graphics, text, natural or synthetic audio, and even face or body animations. MPEG-4 has defined a number of representation tools to address the coding needs of a large variety of media. Coding of such objects, however, is only the first step into constructing a complete multimedia scene. Additional information is needed in order to: 1) describe how these objects should be placed in space and time, 2) how they may interact with each other and the end-user, 3) how to multiplex all this information into one or more streams for delivery over a variety of networks, and 4) ensure proper synchronization among the various streams. This information is the realm of the MPEG-4 Systems specification (Part 1 of the MPEG-4 standard), which is also responsible for the overall architectural definition of MPEG-4. We describe the key features of the MPEG-4 Systems (ISO 14496-1) specification and their encoding methodology: scene description and BIFS, animation streams, object descriptors, the sync layer, object content information, as well as multiplexing. We also describe currently available products using MPEG-4, as well as early product prototypes still in the development stage.

The Java based application engine defined in the Systems part (Part 1) of the MPEG-4 standard is called MPEG-J. MPEG-J, which stands for MPEG-Java, is designed to be used with the parametric, scene description based, presentation engine defined by the same standard. It facilitates programmatic control as opposed to the parametric control provided by the presentation engine by defining a set of Java APIs to access and control the underlying MPEG-4 terminal. An MPEG-J application (MPEGlet) which uses these APIs can be used in the content as an MPEG-J elementary stream similar to audio and video streams. A delivery mechanism is defined in MPEG-J to carry the Java byte code and Java objects that makes an MPEG-J application. Complex algorithmic control can be embedded in such an application to include behavior based on time-varying terminal conditions. The lifecycle and scope of an MPEG-J application is defined in the standard along with a strict security model. Content creation in MPEG-J involves a paradigm shift as it blurs the boundaries between media content and computer programs. As the MPEG-J application can leverage the power of Java, the application scenarios are limited only by the ingenuity of the content creator.

A growing diversity of pervasive devices such as wireless phones and hand-held computers is gaining access to the Internet and other information sources. However, much of the rich multimedia content cannot be easily handled by the terminal devices with limited communication, processing, storage and display capabilities. One of the driving applications of the MPEG-7 multimedia content description standard involves the concept of "Universal Multimedia Access." The basic idea of Universal Multimedia Access (UMA) is the allowance of scalable or adaptive delivery of multimedia to users and terminals regardless of conditions of communication bandwidth or capabilities of terminal devices and their support for media formats. In this paper, we describe how MPEG-7 enables UMA by providing tools for specification of Transcoding Hints, Content Variations, Space and Frequency Views, and Summaries.

The ongoing MPEG-7 standard intends to provide a "Multimedia Content Description Interface." In other words, it will provide a rich set of tools to describe content with a view to facilitating applications such as content based querying, browsing and searching of multimedia content. The MPEG-4 standard provides tools for compressing multimedia content at bitrates that are feasible with typical internet connections. Such bitrates fall significantly short of those supported by prior standards such as MPEG-i and MPEG-2. Thus, in this paper, we present a remote video and still image browsing system that uses MPEG-7 for the querying/browsing/searching and MPEG-4 for compressing any transmitted content. We use descriptors of features such as color, shape and motion to annotate the stored content with MPEG-7 like metadata. The aforementioned descriptors stem from our previous work and are currently in the working draft of the MPEG-7 standard. In our previous work, we have shown the efficacy of each of the descriptors individually. In this paper, we show how we combine some of the features to effectively browse remote video and still image content. Our emphasis is on accurate and quick browsing of the remote content. Our system consists of a video web server with stored MPEG-4 video/still content that is able to support remote requests through a simple browser interface. We have used a combination of cgi script and servletapplet based configurations. We will present a demonstration of our system at the conference. We have already successfully demonstrated it to the Japanese press.

MPEG-7 is a standard for describing audio-visual data. How are previous MPEG standards and applications different from this new MPEG-7 development activity? MPEG-1,2 were primarily standards for compressing audio-visual (AV) data; MPEG-4 was designed to allow manipulation of audio-visual data by converting AV composite data into AV objects that can be composed and accessed within an AV scene. MPEG-7, in fact, is a paradigm shift from the past, a pioneering work that aims to dramatically improve search and retrieval of information over the internet using the inherent semantic and structural features of AV data. It is a multi-disciplinary activity that needs an understanding of low-level structural AV data to high-level semantic meanings of AV data and how these data relate to each other. It leverages on the results of other data representation standards such as SMPTE and Dublin Core but focuses on rich and detailed descriptions of AV data. MPEG-7 compels us to examine and record how we query and browse for information. This paper highlights the nature and scope of the MPEG-7 standard and provides examples of applications and demos currently being developed throughout the world.

With the increase in the amount of digital content that consumers have to deal with, both broadcast content and home captured video and images, there is a real need for tools that help manage multimedia content. In this paper we will describe a system that we are building to manage digital content at home. The system is designed for managing both still images and collections of video, where video can be either video captured by the consumer or clips that are captured from TV broadcast.

MPEG-7 is a standard to provide a standardized way of describing contents of multimedia information. The development of technologies related to the MPEG-7 and exploiting MPEG-7 applications have been a main research interest of us for the last two years. TV broadcasting is one specific area of interest among various applications to which MPEG-7 can be of benefit. The VideoGadgetTM, which is an indexing and retrieval engine of multimedia data, is designed to aid user friendly environment of Digital TV as well as conventional analogue TV. By enabling automatic feature extraction and indexing of received TV program with help of low-cost random access storage device, a dull and conventional TV can turn into an interactive personalized entertainment center. With various user-friendly functionalities such as non-linear browsing of received programs, structure-based navigation, and searching/filtering of programs, the user can view the received program anytime the user prefers in the way the user prefers. Even though the Video-GadgetTM project, which is currently in progress, aims an environment of the Digital TV broadcast with MPEG-7 descriptions, a version of the Video-GadgetTM is under development to actually generate MPEG-7-like descriptions while the analogue stream is received. The real-time generation of the content description in detail brings many problems to solve such as limited I/O bandwidth and computing power and extensive sharing of the system resources. Due to the limited resource, some features are impossible to be extract in the real time, the functionality using those features should be redesigned to use simpler features. Once the Digital TV broadcast with MPEG-7 descriptions become available, the feature extraction part of the Video-GadgetTM will be employed by the MPEG7 description generation server or the service provider and only the retrieval engine including search, browsing, and filtering functionality will remain in the client part.

Along with increased demand for access to distributed information and communication via the Internet, wireless PDA devices and cellular phones, there is a tremendous need for information providers to adapt to various user demands, not only in terms of the diversity of content, but also in how that content is delivered. The diversity of user, network and client capabilities often make traditional approaches such as the storage of multiple representations impractical. In the case of video, one of the primary considerations is the ability to provide content over low bandwidth and noisy networks as well as higher capacity networks. In this paper, we focus on novel techniques to adapt video for its delivery over IP. We provide insight into the existing problems and provide an overview of how our video analysis work is being applied in these areas.

Internet multimedia applications have different requirements than do traditional text-based applications, placing new demands on TCP and UDP, the dc-facto Internet transport protocols. We propose a Selective Retransmission Protocol (SRP) to balance the potentially high loss found in UDP with the potentially high latency found in TCP. SRP uses an applicationspecific decision algorithm to determine whether or not to ask for a retransmission for a lost packet, adjusting the loss and latency to the optimum level for the application. We develop and experimentally evaluate an audioconference using SRP on a wide-area network testhed. We find SRP outperforms both TCP and UDP in terms of multimedia application quality.

For Internet audio applications, much effort has been involved in packet-level error control and delay jitter concealment. In this paper, a packet-based time-scale modification scheme for speech signal is applied to provide adaptive delay concealment at the receiver of an Internet voice session. The adaptive playout algorithm strives to minimize receiver packet droppings for late-arrival packets and premature packets while keeping the end-to-end delay constrained. By stretching the voice segment up/down and incorporating the silence interval, the proposed algorithm could adapt quickly to accommodate fluctuating delays including delay spikes. The evaluation verifies the performance of the proposed adaptive playout, which improves the received speech intelligence under a tightly bounded average playout delay.

This paper describes an implementation of a Delivery Multimedia Integration Framework (DMIF) protocol stack over IP networks with RSVP signaling, and reports the most relevant experimental results. Particular attention is given to a video-on-demand service scenario where a server delivers MPEG-4 multimedia objects, each one composed by a video stream and an audio stream. Two key aspects are investigated. The first aspect is the mapping strategy of QoS DMIF parameters into QoS RSVP parameters. A mapping method is proposed that requires pre-processing of stored media streams. The second aspect is to compare the network resources utilization in two distinct options. The first option is to choose between joining the video and audio streams in a single RSVP reservation or sending each stream in its own RSVP reservation. The second option is to choose between packing data of each stream before sending to network or sending each data packet individually. The DMIF protocol stack was developed for Windows 2000 OS using WinSock 2 and its QoS SP, which supports RSVP. Experimental results are based on MPEG-4 objects composed by H.263 video streams and G.723 audio streams.

Current video coding techniques have various features to enable robust transmission of compressed video data over noisy channels. Among them, the use of resynchronization markers is an essential way to prevent spatial error propagation. In this paper, we propose an optimal approach for placement of resynchronization markers for video communication over noisy channels. Given the number of resynchronization markers assigned to the current frame, we optimally decide the position of each marker in a rate-distortion sense. The proposed method takes into account the channel condition and the expected distortion at the decoder, and incorporates these informations into a rate-distortion framework. Some assumptions and the Viterbi algorithm are employed in order to reduce the complexity of the rate-distortion problem. Various simulation results are shown to justify the performance improvement achieved by the proposed method.

Supporting real-time multimedia applications over the packet wireless network poses a huge challenge due to the stringent QoS requirements and the time-varying and location-varying characteristics of the wireless channel. In this paper, we study a downlink packet scheduling algorithm designed to support real-time applications with hard deadlines over the wireless network. The scheduling depends on the channel characteristics as well as the deadline of each packet. A hybrid TDMA/CDMA physical layer is used to assess the channel condition and to support the target BER of different users. RTP packet headers are used to calculate the deadline of each packet. A modified form of EDF is used to schedule packets in order of urgency and channel condition. Our results indicate that by jointly considering the channel information and the application-level information, a significant performance gain can be achieved. In addition, a more fair service is achieved even though channel conditions of different users vary greatly due to propagation losses and fading effects of the wireless channel.

"Accessing information anytime, anywhere, with any device" is a widely used slogan. The problem, however, is how to bring multimedia services to terminals with limited capabilities. Our approach offers a solution for delivering multimedia presentations to terminals connected to mobile or fixed networks with varying capabilities. Our media wrapper is used to adapt the original presentation according to end-to-end environment characteristics; e.g., the user, terminal, and network type. This paper presents a technique called 'media wrapper' that controls and models the media adaptation process. Using this approach, it is possible to scale and convert multimedia presentations to suit the end user's terminal by making only one original presentation and letting the media wrapper to scale it for the different environments. The adaptation is performed using special algorithms that are defined by dedicated software agents. The algorithms are implemented as agents and are controlled by the media wrapper management module. The agent control is managed based on knowledge base information that contains, e.g., logical rules consisting of the combinations of the needed adaptation parameters. The media wrapper uses CC/PP profiling for defining the parameters that affect the adaptation. These parameters concern terminal, network, and client behavior.

Effective transmission of multiple video signals over a CDMA system simultaneously is investigated in this work. A channel code assignment scheme that efficiently protects compressed bitstreams while minimizing multiple access interference (MAI) is proposed. First, each video signal is coded with a two-layer structure that consists of the base and the enhanced bitstreams according to the bit importance. Then, these bitstreams are protected against transmission errors with channel codes such as RCPC. Better protection of higher bit-rate video requires more multicodes in spreading, which can lead to a severe multiple access interference problem. We set up a framework to a joint design of channel codes and spreading codes with the feedback of the channel status, and provide a solution to deal with the trade-off between channel coding rates and the assigned number of multicodes to achieve efficient transmission. Preliminary experimental results are presented to demonstrate the performance of the proposed channel code assignment scheme.

An adaptive system to support multimedia applications in a wireless network environment is investigated in this research. The proposed system is hierarchical in nature with a cluster of mobile end-hosts connected to a base station, and base stations are connected to a supervisory node, which is in turn connected to a wired infrastructure. Based on our previous experience, the system is designed with an improved service model and a novel call admission scheme that enables more efficient radio channel usage. The service model takes into account both user mobility and the nature of multimedia traffic. Specifically, delay- and/or rate-adaptive services are explored to increase the network performance and utilization under highly variable network connectivity and transmission capacity. The new call admission control scheme is developed under a hybrid call admission control framework. Together with an adaptive resource reservation scheme, it can be adapted to the network situation as well as application requirements. Simulation studies are conducted to show the performance improvement of the proposed scheme.

A dynamic call admission control (CAC) and its associated resource reservation (RR) schemes are proposed in this research based on the guard channel (GC) concept for a wireless cellular system supporting multiple QoS classes. A comprehensive service model is developed, which includes not only mobile terminals' bandwidth requirements but also their different levels of priorities, rate adaptivity and mobility. The proposed CAC policy selects the resource access thresold according to the estimated number of incoming call requests of different QoS classes. The amount of resources to be reserved is dynamically adjusted by considering neighboring-cell higher-priority calls which are likely to handoff. The handoff interaction between adjacent cells is estimated by using radio propagation in terms of the signal-to-noise ratio (SNR) and the distance of each active call in neighboring cells. Experiments are conducted by using the OPNET simulator to study the performance of the proposed scheme under various traffic conditions. It is shown that better QoS guarantees can be provided by the proposed CAC and RR schemes.

In this paper, we describe a visual recognition system of the audience's reactions for multi-point tel-conferences. The system, which works in real-time, can recognize and evaluate some reactions of the audience. It is mainly composed of three subsystems. The first subsystem is a detection and tracking system which is used to detect face and hands from a complex background and track them. The second subsystem is a recognition system which can recognize gestures and poses of the human head, "ask" actions, and the concentration of the audience. The third subsystem is an information fusion system which is used to fuse the information from different sites and evaluate the reactions of the whole audience by using the recognition results in the second system. The outputs of this system are transmitted to the speaker in a remote location so that the speaker can do something according to the information received

Continuous distributed interactive media applications require QoS-capable middleware platforms, in order to meet stringent requirements in terms of losses, bandwidth and jitter. Existing middleware platforms, such as OMG Common Object Request Broker Architecture (CORBA) do not meet these requirements. The major part of the this paper is dedicated to the proposal, presentation and discussion of a middleware platform - named Augmented Reliable Multicast CORBA Event Service, ARMS - developed with the aim to provide quality of service mechanisms for continuous distributed interactive medium support.

Learning on the Internet becomes the main stream in the deployment of distance education and training. In such an environment, both learners and teachers or peer-to-peer are able to join the learning activities synchronously as well as asynchronously. However, the present user interfaces still do not provide enough flexibility and interactively to both learners and teachers. As a result, the users may lack of interests to continuously use the prepared learning environment. In this paper, we design and implement a multimedia user interface, which contains video, audio, text contents, and authoring tools. We integrate the state-of-the-art technologies, such as image segmentation, object tracking, and voice over IP. In the video portion, the system is able to automatically track the teacher movement. Teachers have high degree of freedom in presenting themselves to attract learners' attention. To achieve the above, the developed region-based image segmentation scheme and tracking scheme is realized in real-time. Special voice transmission scheme based on forward error correction is also designed to wrestle with the transmission difficulties on the present Internet. Authoring tools are provided to teachers and learners to improve the convenient utilization of the system.

The release of WebTV was the start of a revolution in the web suffering. Instead of using both PC and television, many technology analysts suggest the integrated unit that will have the widespread availability and easy-of-use of television, combined with the interactive power and flexibility of a PC. In this issue, ATVEF specification is one of the most promising standards in the enhanced television field. In this paper, I would like to propose the prototype of digital data broadcasting system based on the ATVEF specification. This prototype system employs data piping as a mechanism for the delivery of arbitrary user-defined data inserted directly into the payload part of the MPEG-2 Transport Stream packets. This data type includes URL or HTML content based on the ATVEF specification. After the contents feed into the MPEG-2 Transport Stream, it can be delivered to the DTV set-top receiver through the network. Trigger objects received in real-time during the TV broadcast will be used to initiate display or switch content. This prototype system can be used as an example on how to achieve common design goals and how to integrate Digital TV with Web pages based on the ATVEF specification.

The Rate Control phase in MPEG-2 is crucial for the encoding of video sequences for two reasons. First, for timely delivery of video without buffer overflows or underfiows and second for determining indirectly the encoded video quality through moderation ofthe quantization parameter on a macro-block basis. We propose a novel robust estimate which combines local activity estimates with the average activity ofthe previously encoded frame for improving the rate distortion performance of MPEG-2. We then propose a family of exponential modulators for reducing the "over-normalization" effect which occurs when the activity of the macro-block to be encoded is higher than the activity of the previously encoded frame. Extensive experiments show that the proposed low complexity schemes outperform MPEG-2 in terms of PSNR values for the same number of bits produced. We report increases up to 5 db for the luminance component and up to 3.5 db and 3db for the chrominance components respectively

This paper introduces a multi-differential video coding algorithm for the efficient transmission of video conference streams. Because the camera is typically stationary during a video conference, we can take advantage of this information to aid in the robust transmission of the video. Specifically, we describe two simple techniques that can be used to segment the video in to background and foreground information. Using this information, we can then efficiently transmit the video stream while reducing the burstiness seen at the network layer. This technique does not require any changes to the video compression standard (typically H.261 or H.263 for video conferencing) but can reduce the bandwidth required for transmission by 20%. Experimental results are shown with actual video stream traces.

This paper proposes a common designing of a portable video transmitter that is featured by high-quality video compression and high-reliability/efficiency transmission protocol. For flexible video compression on a very small platform, a new software codec is deployed, which is accelerated with an innovative motion vector search algorithm and an adaptive picture prediction type selection. The compressed video is transmitted by the store-and-forward operation fortified with a newly proposed protocol that can deal with the network errors, unpredicted line cut-off and throughput declination in a high delay environment. A prototype system developed based on the proposed architecture is finally described. A notebook-based platform achieves MPEG-2 video transmission even from outdoor environment in particular for an urgent video coverage or quick video reporting.

In movie transmission, video frames are subject to loss due to noise and/or congestion. The loss of video frames could cause a loss of synchronization between the audio and video streams. If not corrected, this cumulative loss can seriously degrade the motion picture's quality beyond viewers' tolerance. Furthermore, many Internet applications providing video and audio streams have to operate in a limited resource environment. The luxury of time and abundant resources is not present, especially for applications like video conferencing and movie provision on demand. In this paper, we develop and study real time motion-based and interpolation-based techniques for performing synchronization-sensitive frame estimation using existing received frames only, without the need for retransmissions or error control information. The estimated frames are then injected at their appropriate locations in the movie stream in order to bring back the synchronization within the tolerance level of viewers. We attempt to find a very close estimation to the original frames at a suitable computation cost. We utilize a normalized luminance model for tracking motion between received frames in the presence of lost data. We ultimately use the generated motion vectors to reconstruct missing data in frames, and to restore synchronization loss between transmitted audio and video streams. We also develop quadratic and linear interpolations for the same purpose. In addition, hybrids of such techniques were developed to best utilize the performance of each developed frame estimation technique under varying degrees of losses and different scene types. Using our techniques, we were able to fully restore synchronization between audio and video streams, with a corresponding estimation of data loss. Both objective and subjective evaluations were performed on the achieved results.

A new technique based on motion compensated prediction is investigated in this research to enhance the quality of interpolated frames in frame-rate up-conversion applications. To playback temporally subsampled video at the full frame-rate, motion compensated interpolation (MCI) is usually employed to reconstruct skipped frames by referencing the decoded frames and motion vectors between them. Since the conventional motion estimation scheme (as adopted in H.26X and MPEG-X) adopts the block matching algorithm (BMA) only between reference and predicted frames, the derived motion vectors for skipped frames are limited in supporting MCI at the decoder end. Thus, by embedding the MCI module in the encoder loop, the performance gain of an embedded encoder improves the frame interpolation accuracy of the decoder. In particular, a two-step approach is considered in this work. First, by incorporating compensation efficiency of both predicted and interpolated frames, a refining scheme for reference motion vectors is proposed. Next, efficient weighting of reference motion vectors, which is similar to window weighting of overlapped block motion compensation (OBMC), is adopted to enhance deformable block-based MCI with four vertex motion vectors. Experimental results are provided to illustrate the performance gain of the proposed frame interpolation scheme.

A content adaptive near lossless image codec is designed in this paper to accommodate a visual perception based rate control scheme. The near lossless codec is a modified version of JPEG-LS, in which the modification is typified by the addition of: (i) two levels of information loss distribution; (ii) line-fitting target prediction; and (iii) visual perception based rate control, implemented via quantization and redistribution of distortion. To minimise the possible excessive information loss being introduced during the rate control process, a technology of so-called moving-across-arrays is proposed to arrange the adjustment of distortion levels. Combined with the line-fitting based target prediction, the proposed scheme successfully controls the compression ratio within a limited range of targets, yet the decoded image quality is maintained competitive to JPEG-LS, the non-controlled one. Extensive experiments are carried out to show that the proposed scheme achieves effective rate control for a range of image samples and competitive image quality in terms of visual perception.

The rapid proliferation of Multimedia content and the increasing number of transactions through the internet has placed high importance to data/,media security. Digital watermarking, a process of embedding secondary digital data in the primary digital medium, is one of the common methods used for copyright protection. In this paper, we propose a technique for robust and invisible watermarking of sprite in MPEG-4. Sprite coding is a new scheme proposed in the MPEG-4 standard for encoding the background in a scene which have less or no local motion. There are three steps involved in the generation of sprites - Initialization, Updating and Blending. The sprite is initialized for the very first Video Object Plane(VOP) of the video object sequence by just placing the VOP at a pre-specified location within the sprite. The sprite is then updated for each of the subsequent VOPs of the video object sequence by estimating the motion of the VOP with respect to the sprite. The VOP is then warped by the estimated motion parameters. The warped VOP and the sprite are then blended together. The blending factor may be different for different pixels on the sprite and this is a function of the number of times this pixel has been blended in the past. The details of the proposed watermarking technique are as follows. The contents of the sprite are first classified as strong texture, smooth and edge regions. The pixels in the sprite are then re-classified depending on the number of time search pixel has been blended. A pseudo noise sequence is generated and this sequence is modulated with the image using the basic sequence amplified according to an apriori classification and pixel blending weights. Abstraction of the watermark is done by demodulating the sprite with the same pseudo-noise sequence. Provisions for extracting the watermark from the individual frames have also been provided. The principal advantage of this watermarking technique is that the watermark bits are modulated onto the sprite depending on the merit of each pi

Two methodologies for efficient audiovisual data encryption are studied in this research, i.e. selective encryption and MHT (multiple-Huffman-table). We first discuss benefits and limitations of selective encryption, and propose a selective encryption scheme for ITU G.723.1 speech coding. A review of all cryptanalysis work of selective encryption algorithms in the literature is given, our cryptanalysis for the JPEG/MPEG sign-bit encryption scheme is presented, and two rules to determine whether selective encryption for a compression system is suitable are concluded. Then, we investigate another encryption methodology via the manipulation of Huffman coding tables of multimedia coding systems. The MHT scheme chooses several different Huffman tables from a vast number of possible candidates, and uses them alternatively to encode multimedia data. The choice of Huffman tables and the order that they are used are kept secret as the key. This method requires very little computational overhead, and can be applied to the encryption of MPEG audio, MPEG video and JPEG/JPEG2000 images.

This paper presents a visible watermarking system using perceptual models. %how and why A watermark image is overlaid translucently onto a primary image, for the purposes of immediate claim of copyright, instantaneous recognition of owner or creator, or deterrence to piracy of digital images or video. %perceptual The watermark is modulated by exploiting combined DCT-domain and DWT-domain perceptual models. % so that the watermark is visually uniform. The resulting watermarked image is visually pleasing and unobtrusive. The location, size and strength of the watermark vary randomly with the underlying image. The randomization makes the automatic removal of the watermark difficult even though the algorithm is known publicly but the key to the random sequence generator. The experiments demonstrate that the watermarked images have pleasant visual effect and strong robustness. The watermarking system can be used in copyright notification and protection.

We propose a content based watermarking technique in multimedia management system. In the proposed technique, a content description technique of MPEG-7 for the multimedia database is adopted into a watermarking technique. With multimedia features described by MPEG-7 standard, we propose a novel watermarking technique where MPEG-7 descriptors are regarded as perceptually significant portions. The watermark is embedded in cooperating with multimedia features such as MPEG-7 descriptor. To verify the feasibility and performance of proposed watermarking technique, experiments with the MPEG-7 database are performed

Typical implementations of standard video compression algorithms use variance-based perceptual masking to adaptively vary the degree of compression noise to video content; the idea being that the human visual system is more sensitive to errors in low variance regions of the image. On a similar note, quality assessment algorithms employ similar spatial-frequency adaptive measures to judge compressed video quality. Such simplistic analysis leads to problems when we encounter "clean edges" in video (for example, text) which although classified as high frequency areas, are still very sensitive to compression noise. Such discrimination between "good" and "bad" texture is particularly important for low bit rate, internet video transmission. In this paper we will propose a novel, computationally simple scheme which further distinguishes between "clean" and "random" edges. By comparing the proposed algorithm with other standard texture classification schemes in literature, we will show that we obtain a substantial improvement in discriminatory ability, albeit at a much lower computational cost. Further more, we will show that the proposed scheme can be applied on a per macro-block basis, thus making it particularly suitable to drive standard, block based video compression algorithms such as H.261/3 or MPEG-1/2 which are quite popular.

To accelerate decoding fractal-coded images, researchers have actively investigated methods for deriving dependency regions. This paper describes a novel approach to expanding dependency regions for fast fractal image decoding. Our approach carefully identifies and encodes data dependency at encoding time in such a way that the decoder guarantees to get the dependency regions and further expands the dependency regions effectively. The experimental results show our approach improves the efficiency of the fractal decoding process significantly.

The algorithms of G.723.1 coder are analyzed and studied in this paper. Searching the large-sized codebook in this coder is still found to be time-consuming. Therefore, in order to speed up the code vector searching and optimize the algorithms, a new fast vector quantization method based on clustering analysis is put forward. The scheme of realizing the specific clustering analysis algorithm is described. Experiments and actual execution on G.723.1 coder show that its executive efficiency has been obviously improved by the optimization while its speech quality has little degradation.

Recognition of patterns irrespective of their actual orientation is approached in a context of estimation theory. In particular, it is shown how using Circular Harmonic Filter banks the problem drastically simplifies from a computational viewpoint. A general purpose pattern detection/estimation scheme is finally illustrated by decomposing the images on an orthogonal basis formed by Laguerre-Gauss Harmonic Functions, and an application example is provided.

In this paper a novel wavelet based robust blind watermarking scheme is presented. Taking into account the property of the human visual system, the proposed scheme embeds the watermark into the wavelet coefficients representative of the image's areas highly dense of details. More specifically a two stages unconventional wavelet decomposition is used. In fact, after having performed a 2D one-level wavelet decomposition, the high frequency wavelet subbands are further wavelet decomposed, thus obtaining those coefficients where the mark is finally embedded. Experimental results demonstrate that this embedding strategy allows to greatly increase the local ratio between the mark and the original image energies, thus leading to robustness even against severe image degradations, without transparency loss.

Most of the watermarking schemes are based on the additive spread-spectrum method since the embedded watermark can resist more serious content distortion and noises. However, they usually fail to be detected when the image undergoes geometrical modifications, such as cropping, rotation, scaling, or even change of the aspect ratio. In this research, we propose a spatial-frequency composite digital image watermarking scheme to make the embedded watermark survive the generalized geometrical transformations. Two watermarks will be embedded, one in the spatial-domain and the other in the frequency-domain of the image. Both embedding processes are based on the spread-spectrum approach. The frequency-domain watermark contains the desired information that will be carried with the host image. The watermark is embedded in the Discrete Fourier Transform (DFT) magnitude so that it is not affected by image translation. The spatial-domain watermarking is used to help recover the image to its original orientation and scale so that the hidden information can then be successfully determined. Experimental results show that the embedded watermark survives generalized geometrical or affine transformations.

Recently, the Video Conference System (VCS), especially desktop VCS becomes one research hot in multimedia technology field. VCS will manage the conference more efficiently with the multimedia database supporting. In this case, we cany out the research of VCS and developed a Video Conference System prototype —DMCS , which is supported by multimedia database with C/S architecture. In this paper, we first introduce the DMCS' architecture, then the solutions to key issues, such as conference model, data flow control policy, conference data index policies etc. At last the testing results are provided.

This article presents a Description Scheme (DS) to describe the audio-visual documents from the video editing work point of view. This DS is based on edition techniques used in the video edition domain. The main objective of this DS is to provide a complete, modular and extensible description of the structure of the video documents based on editing process. This VideoEditing DS is generic in the sense that it may be used in a large number of applications such as video document indexing and analysis, description of Edit Decision List and elaboration of editing patterns. It is based on accurate and complete definitions of shots and transition effects required for video document analysis applications. The VideoEditing DS allows three levels of description : analytic, synthetic and semantic. In the DS, the higher (resp. the lower) is the element of description, the more analytic (resp. synthetic) is the information. %Phil This DS allows describing the editing work made by editing boards, using more detailed descriptors of Shots and Transition DSs. These elements are provided to define editing patterns that allow several possible reconstructions of movies depending on, for example, the target audience. A part of the video description made with this DS may be automatically produced by the video to shots segmentation algorithms (analytic DSs ) or by editing software, at the same time the edition work is made. This DS gives an answer to the needs related to the exchange of editing work descriptions between editing softwares. At the same time, the same DS provide an analytic description of editing work which is complementary to existing standards for Edit Decision Lists like SMPTE or AAF.

With the development of WEB technology, it is more significant to research on prompting the speed of accessing WEB. Several fast accessing methods are introduced and analyzed in this paper. A model WFAM is given here to speed accessing WEB from the view of a designer by O_O technology. WFAM model includes three parts: the first is a mathematical question model for fast Web accessing, the second is three methods which are Accessing Frequency Priority Policy, Accessing Time Priority policy and Colligation Compositor, and the third is corresponding arithmetic for the three methods. It is proved from both theory and experiment to enhance the accessing efficiency on WEB system by WFAM.