In this paper, an image authentication technique by embedding digital "watermarks" into images is proposed. Watermarking a for labeling pictures hiding secret information the images. Sophisticated watermark potential method to discourage unauthorized copying or attest origin of our approach, we embed watermarks with visually recognizable patterns selectively modifying middle-frequency parts image. Several variations proposed are addressed. The experimental results show that successfully...

The objective approaches of 3D image quality assessment play a key role for the development compression standards and various multimedia applications. images faces more new challenges, such as asymmetric stereo compression, depth perception, virtual view synthesis, than its 2D counterparts. In addition, widely used metrics (e.g., PSNR SSIM) cannot be directly applied to deal with these newly introduced challenges. This statement can verified by low correlation between computed measures...

In this paper, a multiresolution-based technique for embedding digital "watermarks" into images is proposed. The watermarking has been proposed as method by hiding secret information in the so to discourage unauthorized copying or attesting origin of images. our method, we take advantage multiresolution signal decomposition. Both watermark and host image are composed representations with different structures then decomposed watermarks resolution embedded corresponding case quality...

Colorization is a computer-assisted process for adding colors to grayscale images or movies. It can be viewed as assigning three-dimensional color vector (YUV RGB) each pixel of image. In previous works, with some hints the resultant chrominance value varies linearly that luminance. However, it easy find existing methods may introduce obvious bleeding, especially, around region boundaries. then needs extra human-assistance fix these artifacts, which limits its practicability. Facing such...

With the idea of social network analysis, we propose a novel way to analyze movie videos from perspective relationships rather than audiovisual features. To appropriately describe role's in movies, devise method quantify relations and construct networks, called RoleNet. Based on RoleNet, are able perform semantic analysis that goes beyond conventional feature-based approaches. In this work, between roles used be context information video scenes, leading corresponding communities can...

The content protection for image-based 3D data is getting more importance with the advance of low cost display devices. depth-image-based rendering (DIBR) image one which consists center and depth generated by provider. left-eye right-eye are rendered from at consumer side. blind watermarking DIBR rarely studied in literature. In this paper, a novel multiple scheme proposed to deal problem images. Besides usual requirement mutual orthogonality among reference patterns watermark embedding, we...

An MPEG-based technique for embedding digital "watermarks" into video is proposed. The watermarking has been proposed as a method to hide secret information in the signals so discourage unauthorized copying or attest origin of media. In method, we take advantage prediction types MPEG bitstreams embed watermarks both intraframe and non-intraframe with different residual masks. experimental results show that an almost invisible difference between watermarked frames original frames, also robust...

The difficulty of vision-based posture estimation is greatly decreased with the aid commercial depth camera, such as Microsoft Kinect. However, there still much to do bridge results human and understanding movements. Human movement assessment an important technique for exercise learning in field healthcare. In this paper, we propose action tutor system which enables user interactively retrieve a exemplar target immediately acquire motion instructions while it front proposed composed two...

Visible watermarking schemes are important intellectual property rights (IPR) protection mechanisms for digital images and videos that have to be released certain purposes but illegal reproductions of them prohibited. techniques protect contents in a more active manner, which is quite different from the invisible techniques. Digital data embedded with visible watermarks will contain recognizable unobtrusive copyright patterns, details host should still exist. The pattern useful scheme...

An image authentication technique by embedding each with a signature so as to discourage unauthorized copying is proposed. The proposed could actually survive several kinds of processing and the JPEG lossy compression.

Semantic context detection is one of the key techniques to facilitate efficient multimedia retrieval. a scene that completely represents meaningful information segment human beings. In this paper, we propose novel hierarchical approach models statistical characteristics several audio events, over time series, accomplish semantic detection. The consists two stages: event and detections. HMMs are used model basic performed in first stage. Then achieved based on Gaussian mixture models, which...

The browsing of quality videos on small hand-held devices is a common scenario in pervasive media environments. In this paper, we propose novel framework for video adaptation based content recomposition. Our objective to provide effective size which emphasize the important aspects scene while faithfully retaining background context. That achieved by explicitly separating manipulation different objects. A generic attention model developed extract user-interest objects, high-level combination...

With the growth of fandom population, a considerable amount broadcast sports videos have been recorded, and lot research has focused on automatically detecting semantic events in recorded video to develop an efficient browsing tool for general viewer. However, professional sportsman or coach wonders about high level semantics different perspective, such as offensive defensive strategy performed by players. Analyzing tactics is much more challenging basketball than other kinds due its...

Location-based service (LBS) is booming up in recent years with the rapid growth of mobile devices and emerging cloud computing paradigm. Among challenges to establish LBS, user privacy issue becomes most important concern. A successful privacy-preserving LBS must be secure provide accurate query [e.g., -nearest neighbor (NN)] results. In this work, we propose a private circular protocol (PCQP) deal accuracy issues LBS. The consists space filling curve public-key homomorphic cryptosystem....

Quick Response (QR) code is a widely used matrix bar with the increasing population of smart phones. However, QR usually consists random textures which are not suitable for incorporating other visual designs (e.g. name card and business advertisement poster). In order to overcome shortcomings noise-like looks codes, we propose systematic beautification framework where appearance composed visually meaningful patterns selected by users, more importantly, correctness message decoding kept...

Out-of-distribution (OOD) detection has seen significant advancements with zero-shot approaches by leveraging the powerful Vision-Language Models (VLMs) such as CLIP. However, prior research works have predominantly focused on enhancing Far-OOD performance, while potentially compromising Near-OOD efficacy, observed from our pilot study. To address this issue, we propose a novel strategy to enhance OOD performances for both and scenarios innovatively harnessing Large Language (LLMs) VLMs. Our...

In the era of deepfakes and AI-generated content, digital image manipulation poses significant challenges to authenticity, creating doubts about credibility images. Traditional forensics techniques often struggle detect sophisticated tampering, passive detection approaches are reactive, verifying authenticity only after counterfeiting occurs. this paper, we propose a novel full-resolution secure learned codec (SLIC) designed proactively prevent by self-destructive artifacts upon...

This paper explores advancements in the Gentry-Sahai-Waters (GSW) fully homomorphic encryption scheme (FHE), addressing challenges related to message data range limitations and ciphertext size constraints. We leverage well-known parallelizing technology—the Chinese Remainder Theorem (CRT)—to tackle decomposition, significantly expanding allowable input entire plaintext space. approach enables unrestricted selection GSW supports parallel operations without intermediate decryption....

We introduce a novel matrix representation for differentially private training and prediction methods tailored to random forest classifiers. Our approach involves representing each root-to-leaf decision path in all trees as row vector matrix. Similarly, inference queries are represented This enables us collectively analyze privacy across multiple queries, resulting optimal DP noise allocation under the Laplace Mechanism. experimental results show significant accuracy improvements of up 40%...

Online photo albums have been prevalent in recent years and resulted more applications developed to provide convenient functionalities for sharing. In this paper, we propose a system named SheepDog automatically add photos into appropriate groups recommend suitable tags users on Flickr. We adopt concept detection predict relevant concepts of probe the issue about training data collection classification. From perspective gathering by web searching, introduce two mechanisms investigate their...

A trusted electronic election system requires that all the involved information must go public. However, it focuses not only on transparency but also privacy issues. In other words, each ballot should be counted anonymously, correctly, and efficiently. this work, an effective e-voting is proposed for voters to minimize their trust in authority or government. We ensure of by putting messages Ethereum blockchain; meantime, individual voter protected via ring signature mechanism. Besides,...

In this letter, we propose a generic and efficient algorithm that can construct both asymmetrical symmetrical reversible variable-length codes (RVLCs). Starting from given Huffman code, the construction is based on two developed codeword selection mechanisms, for case case, respectively; it shown mechanisms possess simple features generate RVLCs easily. addition, new are constructed to be very further reducing coding overheads in MPEG-4 when operating decoding mode.

An automatic facial feature extraction algorithm is presented. The composed of two main stages: the face region estimation stage and stage. In stage, a second-chance growing method adopted to estimate target image. genetic search algorithms are applied extract points within region. It shown by simulation results that proposed can automatically exactly features with limited computational complexity.

This paper presents a new medium, called tiling slideshow, to display photos in tile-like manner, coordinating with the pace of background music. In contrast conventional photo multiple that have similar characteristics are well arranged and displayed at same layout. Motivated by concepts technical writing, each displaying layout is composed larger topic several small-size supportive photos. Based on this idea, proposed slideshow system consists three major components: image clustering,...