The wide application of deep learning techniques has spotlighted the issue of adversarial perturbations. While much attention has been paid to the adversarial and intangible aspects, feasibility is often overlooked, which is an important feature of mimicking real-world knowledge. Traditional attacks on 3D point cloud models, like using global distance loss or tangent plane noise, are time-consuming and generate obvious outliers. Simple loss or metric usage is also inadequate for structured point cloud data. We propose a novel sensing sphere-based sensitivity map. It replicates real-world scenarios and preserves shape, optimizing the efficiency and invisibility of perturbations. By guiding the generation of perturbations on spherical surfaces with explicit constraints and leveraging invertible coordinate transformations for gradient calculation, our method significantly enhances white-box attack stealthiness and black-box query efficiency. Extensive experiments across various point cloud recognition models and against different defense methods validate that our approach achieves outstanding performance and strong resistance.

AI-based automatic aiming cheats (a.k.a., AI aimbots) have proliferated inFirst-Person Shootergames, which grant malicious users an unfair gameplay advantage. Since AI aimbots operate independently of game data and are developed using object detection algorithms, they are difficult to detect with traditional anti-cheating methods. To actively counter AI aimbots, we proposeAdvMap, which introduces invisible adversarial perturbations into game scene elements. In optimizing these adversarial perturbations, we design a mixture-of-misleading loss function that increases the total target confidence score within each misleading bounding box. It mitigates the risk of segment missing even whenAdvMapis obscured, thereby enhancing the robustness. Besides, an L1-norm constraint with a small scale is employed during each update of the adversarial perturbations, which preserves the fidelity of the game scene. In addition, to enable effective adaptation to interact with various elements within game environments, we introduce an image-subspace-based multidirectional optimization strategy. It enables the adversarial perturbations to adaptively fit into each element by leveraging the mapping relationship between the game's 3-D scenes and its corresponding 2-D images. Furthermore, we construct a comprehensive benchmark, which includes various FPS games with different graphics styles and perspectives. Extensive experimental results demonstrate the efficacy of our method in countering various AI aimbot tools on different state-of-the-art object detection methods.

The existing commodity Wi-Fi-based human gait recognition systems mainly focus on a single subject due to the challenges of multisubject walking monitoring. To tackle the problem, we propose Wi-Diag, the first commodity Wi-Fi-based multisubject abnormal gait diagnosis system that leverages only one pair of off-the-shelf commercial Wi-Fi transceivers to separate each subject’s gait information and maintains an excellent performance when the scenario changes. It is an intelligent multisubject gait diagnosis system that can release an experienced doctor from heavy load work. Multisubject abnormal gait diagnosis is modeled as a blind source separation (BSS) issue, and multisubject walking mixed signals are efficiently separated by IC analysis (ICA) approach. This fact is verified by comprehensive theoretical derivation and experimental validation. In addition, CycleGAN is leveraged to mitigate the environmental dependency so that Wi-Diag can be robust when the scenario changes. The excellent performance of Wi-Diag is verified by extensive experiments. The average mean diagnosis accuracy with a maximum group size of four and various scenarios is 87.77%.

Robust reversible watermarking (RRW) is one of the most popular areas in information hiding. Existing schemes have two drawbacks: 1) schemes that can resist conventional attacks often fail to resist geometric attacks, and 2) schemes that can resist geometric attacks often are not robust against conventional attacks and have poor stability. Inspired by the high robustness of fractional-order orthogonal moments (FoOM) and the good feature of resistance to geometric attacks of Zernike moments and pseudo-Zernike moments (ZM/PZM), in this research, FoOM is used to optimize ZM/PZM to obtain FoZM/FoPZM (namely, fractional-order Zernike moments and fractional-order pseudo-Zernike moments). Furthermore, a denoiser is proposed to preprocess the watermarks to improve the robustness against geometric and conventional attacks, the amount of extracted auxiliary information is decreases and the extraction process of auxiliary information is designed to be more stable. Specifically, first, the source of the difference between the watermarked image and the carrier image is identified, that difference is represented with less information, and then that information is embedded into the cover image as auxiliary information. Second, the watermark is embedded in the low-order FoZM/FoPZM component. Finally, the watermarked image is denoised using a denoiser before extracting the watermark. The experimental results show that the scheme has good stability and strong robustness. Compared with existing methods, the proposed scheme has a small and stable auxiliary information size, strong robustness to noise attacks such as Gaussian noise and salt-and-pepper noise attacks, and better resistance to geometric attacks such as rotation and scaling attacks.

As a key task in autonomous driving, 3D object detection based on LiDAR-camera fusion is expected to achieve more robust results by the complementarity of the two sensors. However, LiDAR-camera fusion is non-trivial. An existing problem for this type of detector is that the scale and receptive field of LiDAR point features and image features are not matched, leading to information deficiency or redundancy in fusion. This paper proposes a Point-based Pyramid Attention Fusion (PPAF) module for LiDAR-camera fusion to solve the problem. The PPAF module learns corresponding image features of LiDAR points with a matched scale based on the image feature pyramid and attention mechanism for a better effect of fusion. Furthermore, based on the PPAF module, a new LiDAR-camera fusion-based 3D object detector named FS-Net is proposed, a two-stage detector with LiDAR voxel-based RPN and refinement network based on enriched LiDAR-camera features. Experiments on two public datasets demonstrate the effectiveness of our approach.

The weighted visual cryptographic scheme (WVCS) is a secret sharing technology, where weights are assigned to each shadow (participant) according to its importance. Among WVCS, the random grid-based WVCS (RGWVCS) is a frequently visited subject. It considers the premise of equality of all participants, without taking into account the existence of privileged people in reality. To address this problem of RGWVCS, this paper designs a new model, named as (k, m, n)-RGWVCS (where m < k < n ), in which the secret is encrypted into n shares and sent to k participants. In the recovery end, the secret could be reconstructed by minimum m shares when the privileged join in; otherwise, k shares are needed. The experimental results show that our method has the advantage of no pixel expansion and no codebook design by means of random grid. Moreover, the contrast of our model increased by 32.85% on average compared with that of other WVCS.

Accurate lesion segmentation based on endoscopy images is a fundamental task for the automated diagnosis of gastrointestinal tract (GI Tract) diseases. Previous studies usually use hand-crafted features for representing endoscopy images, while feature definition and lesion segmentation are treated as two standalone tasks. Due to the possible heterogeneity between features and segmentation models, these methods often result in sub-optimal performance. Several fully convolutional networks have been recently developed to jointly perform feature learning and model training for GI Tract disease diagnosis. However, they generally ignore local spatial details of endoscopy images, as down-sampling operations (e.g., pooling and convolutional striding) may result in irreversible loss of image spatial information. To this end, we propose a multi-scale context-guided deep network (MCNet) for end-to-end lesion segmentation of endoscopy images in GI Tract, where both global and local contexts are captured as guidance for model training. Specifically, one global subnetwork is designed to extract the global structure and high-level semantic context of each input image. Then we further design two cascaded local subnetworks based on output feature maps of the global subnetwork, aiming to capture both local appearance information and relatively high-level semantic information in a multi-scale manner. Those feature maps learned by three subnetworks are further fused for the subsequent task of lesion segmentation. We have evaluated the proposed MCNet on 1,310 endoscopy images from the public EndoVis-Ab and CVC-ClinicDB datasets for abnormal segmentation and polyp segmentation, respectively. Experimental results demonstrate that MCNet achieves [Formula: see text] and [Formula: see text] mean intersection over union (mIoU) on two datasets, respectively, outperforming several state-of-the-art approaches in automated lesion segmentation with endoscopy images of GI Tract.

The arrival of cloud computing age makes data outsourcing an important and convenient application. More and more individuals and organizations outsource large amounts of graph data to the cloud computing platform (CCP) for the sake of saving cost. As the server on CCP is not completely honest and trustworthy, the outsourcing graph data are usually encrypted before they are sent to CCP. The optimal route finding on graph data is a popular operation which is frequently used in many fields. The optimal route finding with support for semantic search has stronger query capabilities, and a consumer can use similar words of graph vertices as query terms to implement optimal route finding. Due to encrypting the outsourcing graph data before they are sent to CCP, it is not easy for data customers to manipulate and further use the encrypted graph data. In this paper, we present a solution to execute privacy‐guarding optimal route finding with support for semantic search on the encrypted graph in the cloud computing scenario (PORF). We designed a scheme by building secure query index to implement optimal route finding with support for semantic search based on searchable encryption idea and stemmer mechanism. We give formal security analysis for our scheme. We also analyze the efficiency of our scheme through the experimental evaluation.

Smooth filtering is a common post‐operation which is exploited to blur and conceal the traces of tampered objects. Most of the existing forensic methods aim at detecting only one type of filtering process, such as median filtering or Gaussian filtering, which limits their applications. The authors present a new forensic method based on deep learning technique, which utilises a convolutional neural network (CNN) to automatically learn hierarchical representations from the input images. Unlike conventional CNN models, a modified CNN architecture is specifically designed to identify traces left by the manipulation. A filter layer is added into the CNN. The filtering residual in frequency feature of the input image is extracted by this added layer. The output feature is then fed into the next layer of the CNN. Radon transform is applied to increase the distinctiveness of the residual feature. Experimental results on several public datasets show that the proposed CNN‐based model outperforms some state‐of‐the‐art methods.

Compared to the encrypted-image-based reversible data hiding (EIRDH) method, the encrypted-signals-based reversible data hiding (ESRDH) technique is a novel way to achieve a greater embedding rate and better quality of the decrypted signals. Motivated by ESRDH using signal energy transfer, we propose an improved ESRDH method using code division multiplexing and value expansion. At the beginning, each pixel of the original image is divided into several parts containing a little signal and multiple equal signals. Next, all signals are encrypted by Paillier encryption. And then a large number of secret bits are embedded into the encrypted signals using code division multiplexing and value expansion. Since the sum of elements in any spreading sequence is equal to 0, lossless quality of directly decrypted signals can be achieved using code division multiplexing on the encrypted equal signals. Although the visual quality is reduced, high-capacity data hiding can be accomplished by conducting value expansion on the encrypted little signal. The experimental results show that our method is better than other methods in terms of the embedding rate and average PSNR.

This paper proposes a pseudo 2D string matching (P2SM) technique for high efficiency screen content coding (SCC). The technique uses a primary reference buffer (PRB) and a secondary reference buffer (SRB) for string matching and string copying. In the encoder, optimal reference string searching is performed in both PRB and SRB, and either a PRB or an SRB string is selected as an optimal reference string on a string-by-string basis. If no reference string of at least one pixel is founded for a current pixel, then the current pixel is coded as an unmatched pixel. Compared with HM-16.4${+}$SCM-4.0 reference software, the proposed P2SM technique achieves up to 37.7% Y BD-rate reduction for a screen snapshot of a spreadsheet. On average, using HEVC SCC common test condition and YUV test sequences in text and graphics with motion category, the proposed technique achieves Y BD-rate reduction of 7.7%, 5.0%, 2.6% for all intra (AI), random access (RA) and low-delay B (LB) configurations, respectively in lossy coding with both intra block copy (IBC) and P2SM having the same 4 coding tree units (CTUs) searching range, and bit-rate saving of 6.0%, 3.9%, 3.1% for AI, RA, LB configurations, respectively in lossless coding with IBC having full frame searching range while P2SM having only 2 CTUs searching range, at very low additional encoding and decoding complexity.

As sophisticated photo editing software is increasingly available and the widespread use of multimedia social network, the reliability of digital images becomes more and more important. Photographic splicing, herein defined as a cut-and-paste of image regions from one image onto another image, is difficult to be detected due to the absence of a reference object. To carry out such forensic analysis, we present a novel shadow-based method, with which the fake shadow of the composites can be detected. We show how to estimate the shadow scale factors with a shadow removal technique and, further, how to estimate the growth rate of the penumbra width (GRPW). Inconsistencies in the shadows are then used as evidence of tampering. Compared with other shadow-based forensic methods, the proposed method can not only deal with the problem of shadow cloning in the same image, but also expose the fakery containing the real shadow, which benefit from the estimation of shadow scale factors and GRPW. Comparison results obtained from the splicing forgery detection database verify the ability of our approach.

ABSTRACT To prevent digital image from unauthorized use, image copy detection is an important technique in the field of copyright protection. The conventional methods of image copy detection concentrate on extracting global or local signatures to resist various kinds of copy attacks. However, the global signatures are sensitive to some geometric transformations, such as rotation and cropping, while the local signatures are not discriminative enough to identify copies from similar images. Considering both the robustness and discriminability, a novel image signature based on the combination of global and local signatures is proposed for image copy detection. Firstly, the interest points are detected from a given image by using the Hessian–Affine detector. Secondly, the image is divided into some circle tracks, and thus the interest points are distributed into these tracks. Finally, to combine the advantages of the circle‐track‐based global signature and the interest points, the global distribution characteristics of interest points based on circle tracks are used to generate our image signature. Experimental results demonstrate the effectiveness of our proposed method in the aspects of both robustness and discriminability. Copyright © 2013 John Wiley & Sons, Ltd.

In this paper, we present an arbitrary shape matching (ASM) coding technique for screen contents. In ASM coder, a Coding Unit (CU) is broken into multiple pixel sample strings. Each string called matched string in the CU has a matching string in the previously coded and reconstructed pixel buffer. Then the distance (either 1D or 2D) between the matching string and matched string and the length of the string are entropy-coded into the bitstream buffer. If the length is zero (no matching is found), then the original pixel sample is entropy-coded into the bitstream buffer. Experiments show that for some types of screen contents, ASM can significantly improve coding performance.

Computer screens contain discontinuous-tone content and continuous-tone content. Thus, the most effective way for screen content coding (SCC) is to use two essentially different coders: a dictionary-entropy coder and a traditional hybrid coder. Although screen content is originally in a full-chroma (e.g., YUV444) format, the current method of compression is to first subsample chroma of pictures and then compress pictures using a chroma-subsampled (e.g., YUV420) coder. Using two chroma-subsampled coders cannot achieve high-quality SCC, but using two full-chroma coders is overkill and inefficient for SCC. To solve the dilemma, this paper proposes a mixed chroma sampling-rate approach for SCC. An original full-chroma input macroblock (coding unit) or its prediction residual is chroma-subsampled. One full-chroma base coder and one chroma-subsampled base coder are used simultaneously to code the original and the chroma-subsampled macroblock, respectively. The coder minimizing rate-distortion (R-D) is selected as the final coder for the macroblock. The two base coders are coherently unified and optimized to get the best overall coding performance and share coding components and resources as much as possible. The approach achieves very high visual quality with minimal computing complexity increment for SCC, and has better R-D performance than two full-chroma coders approach, especially in low bitrate.