Contrastive vision-language models like CLIP have achieved impressive results in image-text retrieval by aligning image and text representations in a shared embedding space. However, these models often treat text as flat sequences, limiting their ability to handle complex, compositional, and long-form descriptions. In particular, they fail to capture two essential properties of language: semantic hierarchy, which reflects the multi-level compositional structure of text, and semantic monotonicity, where richer descriptions should result in stronger alignment with visual content. To address these limitations, we propose HiMo-CLIP, a representation-level framework that enhances CLIP-style models without modifying the encoder architecture. HiMo-CLIP introduces two key components: a hierarchical decomposition (HiDe) module that extracts latent semantic components from long-form text via in-batch PCA, enabling flexible, batch-aware alignment across different semantic granularities, and a monotonicity-aware contrastive loss (MoLo) that jointly aligns global and component-level representations, encouraging the model to internalize semantic ordering and alignment strength as a function of textual completeness. These components work together to produce structured, cognitively aligned cross-modal representations. Experiments on multiple image-text retrieval benchmarks show that HiMo-CLIP consistently outperforms strong baselines, particularly under long or compositional descriptions.

Glitch tokens—inputs that trigger unpredictable or anomalous behavior in Large Language Models (LLMs)—pose significant challenges to model reliability and safety. Existing detection methods primarily rely on heuristic embedding patterns or statistical anomalies within internal representations, limiting their generalizability across different model architectures and potentially missing anomalies that deviate from observed patterns. We introduce GlitchMiner, an behavior-driven framework designed to identify glitch tokens by maximizing predictive entropy. Leveraging a gradient-guided local search strategy, GlitchMiner efficiently explores the discrete token space without relying on model-specific heuristics or large-batch sampling. Extensive experiments across ten LLMs from five major model families demonstrate that GlitchMiner consistently outperforms existing approaches in detection accuracy and query efficiency, providing a generalizable and scalable solution for effective glitch token discovery.

In this research, we propose a novel denoising diffusion model based on shortest-path modeling that optimizes residual propagation to enhance both denoising efficiency and quality. Drawing on Denoising Diffusion Implicit Models (DDIM) and insights from graph theory, our model, termed the Shortest Path Diffusion Model (ShortDF), treats the denoising process as a shortest-path problem aimed at minimizing reconstruction error. By optimizing the initial residuals, we improve the efficiency of the reverse diffusion process and the quality of the generated samples. Extensive experiments on multiple standard benchmarks demonstrate that ShortDF significantly reduces diffusion time (or steps) while enhancing the visual fidelity of generated samples compared to prior arts. This work, we suppose, paves the way for interactive diffusion-based applications and establishes a foundation for rapid data generation. Code is available at https://github.com/UnicomAI/ShortDF.

In this paper, a multi-modal model based 3D pop-out video generation framework (CP3) is proposed to solve the shortcomings of the existing video generation technology for accurate control of 3D pop-out effects. 3D pop-out effects create an immersive visual experience by changing the disparity of a particular object so that it appears beyond the screen. However, although software has made some progress in this area, there is currently no effective way to accurately control 3D pop-out effects and generate high-quality video. In addition, the lack of high-quality 3D pop-out effect data sets is also one of the bottlenecks in the field. Therefore, the CP3 framework proposed in this paper utilizes multi-modal models to help 3D video creators make 3D pop-out effects, enhance the audience's sense of immersion and visual comfort, and thus promote the development of 3D effect generation technology. To support the training and evaluation of this framework, a new dataset containing 37000 frames of pop-out effects is constructed, such as text guidance, segmentation results, depth maps, optical flow, and the trajectory of the pop-out target. Through the 3D UNet model based on the potential de-noising diffusion mechanism, combined with the 3D-try module in the CP3 framework and Mask Encoder, this paper has achieved remarkable results in the generation of 3D pop-out effect videos. The results of the experiment show that the CP3 framework demonstrates its advantages in generating immersive 3D pop-out effects in comparison to existing technologies.

We present LeMiCa, a training-free and efficient acceleration framework for diffusion-based video generation. While existing caching strategies primarily focus on reducing local heuristic errors, they often overlook the accumulation of global errors, leading to noticeable content degradation between accelerated and original videos. To address this issue, we formulate cache scheduling as a directed graph with error-weighted edges and introduce a Lexicographic Minimax Path Optimization strategy that explicitly bounds the worst-case path error. This approach substantially improves the consistency of global content and style across generated frames. Extensive experiments on multiple text-to-video benchmarks demonstrate that LeMiCa delivers dual improvements in both inference speed and generation quality. Notably, our method achieves a 2.9× speedup on the Latte model and reaches an LPIPS score of 0.05 on Open-Sora, outperforming prior caching techniques. Importantly, these gains come with minimal perceptual quality degradation, making LeMiCa a robust and generalizable paradigm for accelerating diffusion-based video generation. We believe this approach can serve as a strong foundation for future research on efficient and reliable video synthesis.

Image matching is still challenging in such scenes with large viewpoints or illumination changes or with low textures. In this paper, we propose a Transformer-based pseudo 3D image matching method. It upgrades the 2D features extracted from the source image to 3D features with the help of a reference image and matches to the 2D features extracted from the destination image by the coarse-to-fine 3D matching. Our key discovery is that by introducing the reference image, the source image’s fine points are screened and furtherly their feature descriptors are enriched from 2D to 3D, which improves the match performance with the destination image. Experimental results on multiple datasets show that the proposed method achieves the state-of-the-art on the tasks of homography estimation, pose estimation and visual localization especially in challenging scenes.

It is a challenging task for visually impaired people to perceive their surrounding environment due to the complexity of the natural scenes. Their personal and social activities are thus highly limited. This paper introduces a Large Vision-Language Model(LVLM) based environment perception system which helps them to better understand the surrounding environment, by capturing the current scene they face with a wearable device, and then letting them retrieve the analysis results through the device. The visually impaired people could acquire a global description of the scene by long pressing the screen to activate the LVLM output, retrieve the categories of the objects in the scene resulting from a segmentation model by tapping or swiping the screen, and get a detailed description of the objects they are interested in by double-tapping the screen. To help visually impaired people more accurately perceive the world, this paper proposes incorporating the segmentation result of the RGB image as external knowledge into the input of LVLM to reduce the LVLM’s hallucination. Technical experiments on POPE, MME and LLaVA-QA90 show that the system could provide a more accurate description of the scene compared to Qwen-VL-Chat, exploratory experiments show that the system helps visually impaired people to perceive the surrounding environment effectively.

Anomaly detection without priors of the anomalies is challenging. In the field of unsupervised anomaly detection, traditional auto-encoder (AE) tends to fail based on the assumption that by training only on normal images, the model will not be able to reconstruct abnormal images correctly. On the contrary, we propose a novel patch-wise auto-encoder (Patch AE) framework, which aims at enhancing the reconstruction ability of AE to anomalies instead of weakening it. Each patch of image is reconstructed by corresponding spatially distributed feature vector of the learned feature representation, i.e., patch-wise reconstruction, which ensures anomaly- sensitivity of AE. Our method is simple and efficient. It advances the state-of-the-art performances on Mvtec AD benchmark, which proves the effectiveness of our model. It shows great potential in practical industrial application scenarios.

This paper presents a novel framework for simultaneously implementing localization and segmentation, which are two of the most important vision-based tasks for robotics. While the goals and techniques used for them were considered to be different previously, we show that by making use of the intermediate results of the two modules, their performance can be enhanced at the same time. Our framework is able to handle both the instantaneous motion and long-term changes of instances in localization with the help of the segmentation result, which also benefits from the refined 3D pose information. We conduct experiments on various datasets, and prove that our framework works effectively on improving the precision and robustness of the two tasks and outperforms existing localization and segmentation algorithms.

This paper presents an self-supervised deep learning network for monocular visual inertial odometry (named DeepVIO). DeepVIO provides absolute trajectory estimation by directly merging 2D optical flow feature (OFF) and Inertial Measurement Unit (IMU) data. Specifically, it firstly estimates the depth and dense 3D point cloud of each scene by using stereo sequences, and then obtains 3D geometric constraints including 3D optical flow and 6-DoF pose as supervisory signals. Note that such 3D optical flow shows robustness and accuracy to dynamic objects and textureless environments. In DeepVIO training, 2D optical flow network is constrained by the projection of its corresponding 3D optical flow, and LSTM-style IMU preintegration network and the fusion network are learned by minimizing the loss functions from ego-motion constraints. Furthermore, we employ an IMU status update scheme to improve IMU pose estimation through updating the additional gyroscope and accelerometer bias. The experimental results on KITTI and EuRoC datasets show that DeepVIO outperforms state-of-the-art learning based methods in terms of accuracy and data adaptability. Compared to the traditional methods, DeepVIO reduces the impacts of inaccurate Camera-IMU calibrations, unsynchronized and missing data.

This paper presents a novel system that enables intelligent robots to exhibit realistic body gestures while communicating with humans. The proposed system consists of a listening model and a speaking model used in corresponding conversational phases. Both models are adapted from the sequence-to-sequence (seq2seq) architecture to synthesize body gestures represented by the movements of twelve upper-body keypoints. All the extracted 2D keypoints are firstly 3D-transformed, then rotated and normalized to discard irrelevant information. Substantial videos of human conversations from Youtube are collected and preprocessed to train the listening and speaking models separately, after which the two models are evaluated using metrics of mean squared error (MSE) and cosine similarity on the test dataset. The tuned system is implemented to drive a virtual avatar as well as Pepper, a physical humanoid robot, to demonstrate the improvement on conversational interaction abilities of our method in practice.

Steganalysis is the reasonable method to detect whether the transmitted media content contains secret messages (e.g. business secrecy). This paper proposes two steganalysis methods to estimate the modification ratio of F5 steganography and its improved version that are popularly used to hide secrecy in images. The proposed methods measure the distance between the coefficient histogram of a given image and that of an estimated stego image. The distance is measured based on relative entropy that has the superiority of measuring the distance between two distributions. The estimated modification ratio can be used to distinguish the stego images marked by F5 steganography or its improved version from the original images. Experimental results are given to show that the proposed methods outperform the existing quantitative steganalysis methods against F5 steganography and its improved version.

For hiding messages into multiple least significant bit (MLSB) planes, a new weighted stego-image (WS)\ steganalysis method is proposed to estimate the ratio of messages hidden into each bit plane. First, a new WS with multiple weights is constructed, and it is proved that when the squared Euclidean distance between the WS and the cover image is minimal, the weight parameters are equal to the embedding ratios in MLSB planes. Afterward, based on this result and an estimation of cover image, a simple estimation equation is derived to estimate the embedding ratio in each bit plane. Experimental results show that the new steganalysis method performs more stably with the change of embedding ratios than typical structural steganalysis, and outperforms the typical structural steganalysis method on the estimation accuracy when the embedding ratio in any bit plane is larger than 0.4.

In cloud computing environment, various innovative multimedia (text, image, audio, video, graph, etc.) services appear. For example, users may work on one document or drawing together over Internet, users can push their complicated computing tasks to cloud servers, and users can forward their multimedia content (e.g., TV program) to cloud servers with the aim of obtaining the analyzed result (e.g., a short summarization). To support these services, various computing or management techniques are required, for example, multimedia processing, coding, communication, storage and access. Additionally, in cloud computing, security and privacy become very important issues, for example, user authentication and authorization, information exchange between clients and cloud servers, privacy protection of user-uploaded content, etc. This special issue aims to collect the best papers contributed by leading experts, and expose the readership to the latest research results on multimedia computing and management in cloud computing environment. It includes a number of related topics, demonstrates pioneer work, investigates the novel solutions and discusses the future trends in this field. This issue is composed of five papers carefully selected. Each paper is reviewed by at least two experts, with at least two rounds. The first paper, ‘Multimedia Applications and Security in MapReduce: Opportunities and Challenges’ by Z. Yu et al. 1, reviews cloud-based multimedia applications in MapReduce including texts, images, audios and videos, and investigates the arising security issues and potential countermeasures. Since MapReduce is a popular programming model for distributed storage and computation in the cloud, the paper may provide valuable information to readers. In the second paper, ‘Optimizing Queries with Expensive Video Predicates in Cloud Environment’ by L. Yu et al. 2, the methods for processing video data and executing expensive video predicates in a cloud environment are proposed to solve the real time queries of video surveillance in supermarkets. These methods prove to significantly reduce the amount of video data transmission and the number of expensive video predicates operations. It may be a good example of cloud-based media computing. The third paper, ‘Jump-start Cloud: Efficient Deployment Framework for Large-Scale Cloud’ by Y. Lin et al. 3 designs the jump-start cloud that uses an efficient cloud deployment scheme to minimize the cloud time. It may be beneficial to minimize cloud time for clouds serving multimedia applications, as they occupy much more significant cloud resources, such as I/O and networking bandwidth and computational costs in processing or transcoding multimedia contents. The tests through a Hadoop-based benchmark and MapReduce applications show the designed scheme's competences. In the fourth paper, ‘Cold chain logistics system based on cloud computing’ by X. Chen 4, a cold chain logistics system based on cloud computing is designed. In this system, the cloud computing platform can be used to connect the database between cold chain logistics and external customers, so that each database connection terminal can keep track of and update the data. It thus brings better cooperation between cold chain logistics and their customers, realizes co-control of product sales information, accelerates the speed of cold chain logistics and maximizes the interests of all parties. The fifth paper, ‘On Multi-Watermarking in Cloud Environment’ by J. Wang and S. Lian 5, focuses on the application scenarios of multi-watermarking in cloud environment by investigating the secure media distribution models suitable for cloud-based platform. In detail, the multi-watermarking based on additive rule is studied and then analyzed theoretically, and some experimental results are given to show the performances, such as security, robustness and imperceptibility. It initiates the interesting research topic of media copyright protection in cloud environment. The guest editors wish to thank Prof. Geoffrey Fox for providing the opportunity to edit this special issue on Multimedia Computing and Management in Cloud Environment, and other editors for providing latest publishing information and making this issue published. We would also like to thank the authors for submitting their works as well as the referees who have critically evaluated the papers within the short stipulated time. Finally, we hope the reader will share our joy and find this special issue very useful.

SUMMARY Multiwatermarking is the technique to embed multiple messages into the same media content. It can be used to identify media content's owners or trace the illegal redistributors in media distribution. Especially in cloud environment with multiple content providers or users, multiwatermarking may have potential applications. However, some performances of multiple watermark embedding, for example, the distortion and security, are not considered. In this paper, the application scenarios of multiwatermarking in cloud environment are figured out, which investigate the secure media distribution models suitable for cloud‐based platform. Additionally, the multiwatermarking based on additive rule is studied and then analyzed theoretically from two aspects, one of which is the multi‐embedding of same watermark sequence and the other is the multiembedding of different watermark sequences. The analysis emphasizes on four aspects; that is, robustness, capacity, distortion, and security. And each of which considers two cases, that is, the authorized embedding and unauthorized embedding. Additionally, the correlation detection based on multiwatermarking is proposed, and for each case, the corresponding suggestions to improve system performance are proposed. Some comparative experiments are performed, and results are given to show the validity of these theoretical analysis.Copyright © 2011 John Wiley & Sons, Ltd.

SUMMARY Cloud computing has recently attracted great attention, both commercially and academically. MapReduce is a popular programming model for distributed storage and computation in the cloud. In this paper, we survey cloud‐based multimedia applications, identifying the open issues and challenges which arise when MapReduce is used for cloud computing. Copyright © 2011 John Wiley & Sons, Ltd.

SUMMARY With the rapid development of cloud computing, software applications are shifting onto cloud storage rather than remaining within local networks. Software distributions within the cloud are subject to security breaches, privacy abuses, and access control violations. In this paper, we identify an insider threat to access control which is not completely eliminated by the usual techniques of encryption, cryptographic hashes, and access‐control labels. We address this threat using software watermarking. We evaluate our access‐control scheme within the context of a Collaboration‐oriented Architecture, as defined by The Jericho Forum. Copyright © 2011 John Wiley & Sons, Ltd.

Image authentication is usually approached by checking the preservation of some invariant features, which are expected to be both robust and discriminative so that content-preserving operations are accepted while content-altering manipulations are rejected. However, most of existing features have not obtained convincing performance due to insufficiency of experiments and over biasing of robustness. Motivated by the sparse coding strategy discovered in primary visual cortex, we explore the possibility of using sparse coding coefficients for image authentication. Through extensive experiments, we discover that the proposed feature bears great discrimination as well as robustness, which indicates the effectiveness of sparse coding as a new invariant feature for image authentication.

Multimedia content is a suitable carrier for secret communication. This paper focuses on the steganalysis technique which aims to get the forensic of secrecy existing in multimedia carriers. A key concern for designing a blind steganalysis algorithm is the selection of statistic features. The Probability Density Function (PDF) moment and Characteristic Function (CF) moment are two typical kinds of statistic features commonly used in blind steganalysis. And generally, the features are computed from the subbands of transform domains, such as the wavelet coefficient subbands, the prediction subbands of wavelet coefficients, the prediction error subbands of wavelet coefficients, the wavelet coefficient subbands of image noise, and the log prediction error subbands of wavelet coefficients. To decide which feature is more sensitive to message embedding and useful for steganalysis is important and urgent. Till now, few works have focused on this topic, and they can only give some experimental results without theoretical analysis. Additionally, few frequency subbands have been investigated. To solve this problem, this paper reviews existing feature computing algorithms, compares the two kinds of features, the PDF moments and the CF moments, by analyzing the change trends of the statistic distribution parameters of various frequency subbands before and after message embedding, and so that provides a theoretical basis for the steganalysis feature selection and extraction. These theoretical results are further confirmed by experimental results. This is the first work to provide thorough theoretical analysis on so many feature computing algorithms. It is expected to provide valuable information to researchers or engineers working in the field of steganography forensics or steganalysis.

Some joint fingerprinting and decryption schemes were reported recently for secure multimedia distribution. However, most of them need to be investigated before practical applications. In this paper, the secure distribution scheme proposed by Lemma et al. is investigated and improved. Since this scheme aims to distribute multimedia content by encryption and watermarking, some important performances determine its practicability, including the perceptual security of the encryption operation, the imperceptibility of the embedded watermark and the robustness of the embedded watermark. Some flaws are found in the scheme, such as the low encryption strength, the data overflow caused by encryption/decryption and the low correlation value caused by collusion, which degrade its performances greatly. To improve the scheme, some means are proposed, including media preprocessing, media encryption based on module addition and collusion-resistant fingerprint encoding. Comparative experiments show that better performances are obtained by the improved means. The analysis method proposed in this paper can be used to investigate some other joint fingerprinting and decryption schemes.

The presence of fewer samples and large number of input features increases the complexity of the classifier and degrades the stability. Thus, dimension reduction was always carried before supervised learning algorithms such as neural network. This two-stage framework is somewhat redundant in dimension reduction and network training. This paper proposes a novel one-stage learning algorithm for high-dimension small-sample data, called Projection Vector Machine (PVM), which combines dimension reduction with network training and removes the redundancy. Through dimension reduction operation such as singular vector decomposition (SVD), we not only reduce the dimension but also obtain the size of single-hidden layer feedforward neural network (SLFN) and input weight values simultaneously. This size-fixed network will become linear programming system and thus the output weights can be determined by simple least square method. Unlike traditional backpropagation feedforward neural network (BP), parameters in PVM don't need iterative tuning and thus its training speed is much faster than BP. Unlike extreme learning machine (ELM) proposed by Huang [G.-B. Huang, Q.-Y. Zhu, C.-K. Siew, Extreme learning machine: theory and applications, Neurocomputing 70 (2006) 489-501] which assigns input weights randomly, PVM's input weights are ranked by singular values and select the optimal weights order by singular value. We give proof that PVM is a universal approximator for high-dimension small-sample data. Experimental results show that the proposed one-stage algorithm PVM is faster than two-stage learning approach such as SVD+BP and SVD+ELM.

Ubiquitous multimedia applications are becoming more and more popular. However, the solutions that confirm content and interaction security in these applications are still open issues because of various network convergences and device interconnections. This paper investigates a ubiquitous multimedia scheme and proposes a secure service solution. In the ubiquitous scheme, the multimedia content is encoded with scalable video coding and broadcasted via digital video broadcasting for handheld terminals (DVB-H) to mobile terminals, the access right is transmitted by global system for mobile (GSM) channel, and the media content and access right can also be transmitted from mobile terminals to home TV through WiFi based wireless local area network. The proposed secure solution supports three kinds of business models by using various content encryption modes and secure interaction protocols. The solution's security is evaluated and discussed. Since few works have been done to solve this problem, this paper is expected to attract more researchers.

Collusion is a major menace to image fingerprint. Recently, an idea is introduced for collusion-resilient fingerprint by space desynchronizing including RST (rotation, scaling and translation) and random bending. In this paper, we proposed a new space desynchronized fingerprint method which is for large scale distribution: image is enlarged and encrypted at sender side; at user side, key sequence is used to only decrypt some parts of image and compose these parts to make a new image copy with fingerprint embedded. This method has two priorities: first, only one image is needed to send to or be downloaded by all users with fingerprint embedded; secondly, the embedded fingerprint is robust to collusion attack. For diminishing the degradation introduced by desynchronization, we consider the image content such as human face. Experiments indicate the effectiveness of our method.

Some joint fingerprinting and decryption (JFD) schemes are recently reported. However, most of them need to be investigated before practical applications. In this paper, the audio JFD scheme proposed by Lemma et al. is investigated and evaluated. Some flaws are found in the scheme, such as the low encryption strength, the data overflow caused by encryption and the fragileness to collusion attacks. To improve the scheme, some means are proposed, such as the audio encryption without overflow and the collusion-resistant fingerprint encoding. Comparative experiments show that better performances are obtained by the improved means. The analysis method proposed in this paper can be used to investigate some other JFD schemes.

With the rapid progress in information technology and an enormous amount of media appearing over Internet, e.g., text, audio, speech, music, image, and video, guaranteeing information security is becoming increasingly important. Several pivotal challenges include copyright protection, integrity verification, authentication, and access control etc. As a consequence, the subject of security protection in multimedia communication has attracted intensive research activities in academia, industry, and also government. Nowadays, multimedia data are used more and more widely in human's daily life. The typical applications include audio broadcasting, Digital TV, Mobile TV, etc., which are constructed on multimedia communication techniques. Associated with these applications, there are some security issues, e.g., multimedia content security, payment security and user privacy. To solve these issues, some means are required. During the past decades, various techniques have been reported for secure multimedia communication, including key management, multimedia encryption, authentication, digital watermarking, digital fingerprinting, access control, and digital rights management. Among them, multimedia encryption, authentication, digital watermarking, and digital fingerprinting aim to protect multimedia content's confidentiality, integrity, ownership, and traitor traceability. For example, digital fingerprinting embeds a unique customer code into multimedia content in order to produce a unique copy for the certain customer. Thus, the illegal redistribution of the media copy can be traced by detecting and comparing the embedded customer code. Some other techniques, e.g., key management, access control, and digital rights management, are able to protect payment security and user privacy. Taking access control for example, it permits only the authorized customer to access the multimedia content, while some other customers without payment often have limited access right. Additionally, in different networks such as Internet, 3G wireless, DVB-H, and p2p, different secure protocols and algorithms are required to provide the system security. For example, in DVB-H, the broad cast mode is used for data transmission, while in Internet-based services, the unicast or multicast mode is often used. Different from them, in p2p, the multi-hop transmission without the server is often used. When different networks are converged, the secure system with interactivity and interoperability is expected. All these topics are in active development. Furthermore, devices like digital cameras, mobile/video phones, graphics processing units, DVD player, etc. are expected to be equipped with such security mechanisms. In these situations, software solutions may not be adequate to provide high real-time performance. On the other hand, hardware assisted solutions are much better for easy integration with multimedia hardware, which has low power consumption, higher reliability/availability, and low cost. The aim of this special issue is to present a collection of high-quality research papers that report the latest research advances in secure transmission or distribution of multimedia content than on multimedia content protection. In this special issue, we selected 7 papers, which can demonstrate advanced works in this field. A detailed overview of the selected works is given below. The first paper, A Secure Virtual Point of Service for Purchasing Digital Media Content over 3G Wireless Networks, presents the secure virtual point of service (SVPOS) for secure multimedia content transactions and payments. The proposed solutions based on 3rd generation partnership project (3GPP) generic authentication architecture (GAA) and standard charging and billing protocols (e.g., Parlay-X) guarantees some desirable privacy and security properties, including privacy of subscribers, protection of 3G operator and merchant, and interaction security. The scheme is suitable for cell phone based merchant purchase in 3G networks. The second paper, Realising Time-limitation for Cryptographic Keys in Secure Multimedia Distribution, reports a promising transmission infrastructure and key construction scheme for future multimedia delivery. In particular, the scheme is transmitting the encoded layers on separate streams which will be secured using SRTP. This allows fast adaptation by just dropping layers if necessary. The session keys used for SRTP are time-limited and hierarchically dependent such that only one key is necessary for the client to decrypt all desired layers. Although the prerequisites for choosing the initial values of the time-limited keys are fairly high, this structure implies less administration effort. The third paper, Integrating Fingerprint with Cryptosystem Internet-Based Live Pay-TV System, proposes a new architecture to protect contents from unauthorized viewing and illegal redistribution for an internet-based live pay-TV system. Code embedding is considered in order to enhance the resistance against collusion of fingerprinted video. The paper has demonstrated that the proposed scheme can provide higher security and better trade-off between image and video encryption and fingerprinting imperceptibility than existing works. The paper also indicates the ongoing implementation of the pay broadcasting system based on the proposed architecture. The fourth paper, User Plane Security Alternatives in the 3G Evolved Multimedia Broadcast Multicast Service (e-MBMS), proposes a secure multicast overlay (SMO) approach and describe how to implement based on a proof-of-concept test-bed over public domain Linux routers. The paper demonstrates that SMO is able to achieve a low risk of denial of service attacks, significant advantages in terms of impact on the architecture and on device requirements, and high security association management and key management. The fifth paper, Step-wise Inter-frame Correlation-based Steganalysis System for Video Streams, reports a collusion scheme among video frames to find out whether secret data is hidden in frames or not. The effect of local motion interfering detection precision is studied. The local motion and message embedded in video frame is treated as a bimodal noise. To reliably detect the existence of the embedded message, block-wise correlation-based steganalysis scheme is proposed to reduce the local motion interfering effect. The video steganalysis process is divided into two stages. In the first stage, suspicious video frames will be recognized by decision module employing features extracted with a light-weight collusion scheme. In the second stage, suspicious frames will be analyzed critically by the present powerful image steganalysis schemes. In addition, the determine principle is also studied to reduce the false positive rate in first stage. Experimental results have been presented to show the performance of the proposed strategy. The sixth paper, AuthoCast - a Mobility-compliant Protocol Framework for Multicast Sender Authentication, introduces a protocol framework for authenticating multicast sources and securing their mobility handovers. Using a self-consistent, one-way authentication based on cryptographically generated addresses, a common design is derived to jointly comply with the mobile any source and source specific multicast protocols that are currently proposed. This light-weight scheme smoothly extends the unicast enhanced route optimization for mobile IPv6 and adds only little overhead to multicast packets and protocol operations. The seventh paper, Early Security Key Exchange for Encryption in Mobile IPv6 Handoff, proposes an early security key exchange for encryption in Mobile IPv6 handoff in order to reduce the security latency. In the approach, two issues are addressed in dealing with the latency within the encryption technology during the handover. First, the study extends the early binding update method to deal with the long security exchange negotiation time for the Mobile IPv6 handoff. Secondly, the study adopts the security access gateway (SAG) to solve the limited computing and memory in the mobile node. In conclusion, this issue of Security and Communications Networks offers a ground-breaking view into the recent advances in secure multimedia communications. This issue offers both academic and industry appeal- the former as a basis toward future research directions, and the latter toward viable commercial applications. Finally, we would like to express our gratitude to the Editor-in-Chief, Dr. Hsiao-Hwa Chen for his advice, patience, and encouragements since the beginning until the final stage. Special thanks go to Michelle in Wiley during the production. We thank all anonymous reviewers who spent much of their precious time reviewing all the papers. Their timely reviews and comments greatly helped us select the best papers in this special issue. We also thank all authors who have submitted their papers for consideration for this issue. We hope you will enjoy reading the great selection of papers in this issue.

In this paper, a secure multimedia distribution scheme resistant to collusion attacks is proposed. In this scheme, the multimedia content is modulated bynpseudorandom sequences at the server side, which generates the unintelligible multimedia content, and then demodulated under the control of the fingerprint code at the customer side, which produces the multimedia content contains a unique code. The demodulation process adopts collusion-resistant fingerprint codes to determine which sequences will be removed from the received multimedia content. Since the collusion-resistant fingerprint code is used, the colluders who combine different copies together can be detected. Compared with existing schemes, the collusion-resistant code is used in the proposed scheme, which confirms the robustness against collusion attacks. This scheme provides a good choice for secure multimedia content distribution.

A secure video distribution scheme is proposed, which embeds a fingerprint code into the video content during decryption process. At the server side, the video content is scrambled by motion vector (MV) encryption. At the customer side, the video content is decrypted and fingerprinted simultaneously under the control of both the key and the fingerprint. For MV decryption and fingerprint embedding are both based on MV modification, they are combined into homogenous operations. Thus, it is difficult for attackers to get the clear video content from the gap between the decryption operation and the embedding operation. To counter collusion attacks, the fingerprint can be encoded with collusion-resistant codes before being embedded. Furthermore, by improving the watermarking strength, the colluded copy's quality will be reduced greatly, which makes collusion attacks out of work.

Quantization-embedder (QE) is the basic embedding manner, which is composed of partition function, quantization function, conversion function and encoding function. For achieving statistical security, a steganographcial method based on combination of QEs is proposed to preserve the distribution of stego signal toward that of cover signal, and a practical algorithm is implemented in prediction-error domain. Experimental results show that the proposed method has high preservation ability to keep statistical security.

A scheme is proposed to implement commutative video encryption and watermarking during advanced video coding process. In H.264/AVC compression, the intra-prediction mode, motion vector difference and discrete cosine transform (DCT) coefficients' signs are encrypted, while DCT coefficients' amplitudes are watermarked adaptively. To avoid that the watermarking operation affects the decryption operation, a traditional watermarking algorithm is modified. The encryption and watermarking operations are commutative. Thus, the watermark can be extracted from the encrypted videos, and the encrypted videos can be re-watermarked. This scheme embeds the watermark without exposing video content's confidentiality, and provides a solution for signal processing in encrypted domain. Additionally, it increases the operation efficiency, since the encrypted video can be watermarked without decryption. These properties make the scheme a good choice for secure media transmission or distribution

A secure distribution scheme is proposed for compressed videos, which is based on a compression domain watermarking algorithm. At the server end, the watermarking-based encryption algorithm is used to encrypt the compressed data stream. At the receiver end, the joint fingerprint embedding and decryption (JFD) algorithm is adopted to decrypt and fingerprint video data simultaneously. For the features of the watermarking algorithm, this scheme is easy to extract the fingerprint and detect the colluders, and is suitable for distributing compressed videos. Theoretical analysis and experimental results prove its practice.

Chaos and neural networks have both been used in data encryption because of their cipher-suitable properties, such as parameter-sensitivity, time-varying, random-similarity, etc. Based on chaotic neural networks, a hash function is constructed, which makes use of neural networks' diffusion property and chaos' confusion property. This function encodes the plaintext of arbitrary length into the hash value of fixed length (typically, 128-bit, 256-bit or 512-bit). Its security against statistical attack, birthday attack and meet-in-the-middle attack is analyzed in detail. Its properties make it a suitable choice for data authentication

A novel steganographic method based on non-uniform generalized LSB is proposed. Different from the traditional distortion measure MSE and PSNR, the structural-similarity based image quality assessment is used to measure the distortion caused by the data hiding. With the given maximum allowable distortion, dynamic programming is performed to find the optimum substitution depth vector to achieve the maximum capacity. Experiments show the proposed method can achieve higher embedding payload while keeping smaller distortion.

Video encryption is a suitable method to protect video data. There are some disadvantages in the algorithms proposed before. In this paper, we propose a novel fast MPEG video encryption algorithm, which encrypts run length codes with chaotic run-length encryption algorithm (CREA), encrypts the signs of motion vectors with security-enhanced chaotic stream cipher (SECSC) and distributes keys with chaotic key distributor (CKD) at the same time. Its security, compression ratio and computational complexity are analyzed in details. Experimental results show that, the algorithm has less effect on compression ratio than the algorithms confusing DCT coefficients do, does not change file format, and is of low cost. Thus, it is suitable for secure video encoding with real-time requirement.

Perceptual cryptography encrypts multimedia data in the compression domain and causes some degradation to the decoded data. If customers are interested in the multimedia data, they can pay for one of higher quality. A perceptual cryptography on SPIHT encoded images or videos is presented. By confusing different numbers of wavelet coefficients and encrypting different numbers of coefficients' signs, the images or videos can be degraded to different degrees. The encryption strength can be adjusted according to a certain quality factor, the technique supports direct bit-rate control, and is of low cost, which makes it suitable for multimedia applications with a real-time operating requirement, such as video conferencing, multimedia networks, mobile multimedia, and so on.

Image encryption is a suitable method to protect image data. The encryption algorithms based on position confusion and pixel substitution change compression ratio greatly. In this paper, an image encryption algorithm combining with JPEG encoding is proposed. In luminance and chrominance plane, the DCT blocks are confused by pseudo-random SFCs (space filling curves). In each DCT block, DCT coefficients are confused according to different frequency bands and their signs are encrypted by a chaotic stream cipher. The security of the cryptosystem against brute-force attack and known-plaintext attack is also analyzed. Experimental results show that, the algorithm is of high security and low cost. What's more, it supports direct bit-rate control or recompression, which means that the encrypted image can still be decrypted correctly even if its compression ratio has been changed. These advantages make it suitable for image transmission over network.