Most deep neural network (DNN) based image watermarking models often employ the encoder-noise-decoder structure, in which watermark is simply duplicated for expansion and then directly fused with image features to produce the encoded image. However, simple duplication will generate watermark over-redundancies, and the communication between the cover image and watermark in different domains is lacking in image feature extraction and direction fusion, which degrades the watermarking performance. To solve those drawbacks, this paper proposes a Transformer-based soft fusion model for robust image watermarking, namely WFormer. Specifically, to expand watermark effectively, a watermark preprocess module (WPM) is designed with Transformers to extract valid and expanded watermark features by computing its self-attention. Then, to replace direct fusion, a soft fusion module (SFM) is deployed to integrate Transformers into image fusion with watermark by mining their long-range correlations. Precisely, self-attention is computed to extract their own latent features, and meanwhile, cross-attention is learned for bridging their gap to embed watermark effectively. In addition, a feature enhancement module (FEM) builds communication between the cover image and watermark by capturing their cross-feature dependencies, which tunes image features in accordance with watermark features for better fusion. Experimental results show that the proposed WFormer outperforms the existing state-of-the-art watermarking models in terms of invisibility, robustness, and embedding capacity. Furthermore, ablation results prove the effectiveness of the WPM, the FEM, and the SFM.

Multi-image steganography ensures privacy protection while avoiding suspicion from third parties by embedding multiple secret images within a cover image. However, existing multi-image steganographic methods fail to model global spatial correlations to reduce image damage at the low computation cost. Moreover, they do not account for the anti-distortion capability of the cover image, which is crucial for achieving imperceptible and ensuring security. To overcome these limitations, we propose StegMamba, a distortion-free immune-cover for multi-image steganography architecture with a state space model. Specifically, we first explore the potential of the linear computational cost model Mamba for data hiding tasks through a steganography Mamba block (SMB), whose efficiency makes it suitable for real-time applications. Subsequently, considering that images with distortion resistance reduce embedding damage, the original cover image is reconstructed through immune-cover construction module (ICCM) and associated with the steganography task. Moreover, well-coupled features facilitate fusion, and thus a wavelet-based interaction module (WIM) is designed for effective communication between the immune-cover and the secret images. Compared with the state-of-the-art global attention-based methods, the proposed StegMamba obtains PSNR gains of 3.30 dB, 1.37 dB, and 1.92 dB for the stego image, and two secret recovery images, respectively, and the reduction of 2.87% in detection accuracy for anti-steganalysis. This code is available at https://github.com/YuhangZhouCJY/StegMamba.

As a popular multimedia representation, 3D Point Clouds (PC) inevitably encounter distortion during their acquisition, processing, coding, and transmission, resulting in visual quality degradation. Therefore, it is critical to propose a Point Cloud Quality Assessment (PCQA) method to perceive the visual quality of PC. In this paper, we propose a no-reference PCQA method through structure sampling and clustering based on graph, which consists of two-stage pre-processing, quality feature extraction, attention-based feature fusion, and feature regression. For pre-processing, considering the Human Visual System (HVS) tendency to perceive distortions in both the global structure and local details of PCs, a two-stage sampling strategy is introduced. Specifically, to adapt to the irregular structure of PCs, it introduces structural key point sampling and local cluster to capture both global and local information, respectively, thereby facilitating more effective learning of distortion features. Then, in quality feature extraction, two modules are designed based on the two-stage pre-processing results (i.e., Global Feature Extraction (GFE) and Local Feature Extraction (LFE)) to respectively extract global and local quality features. Additionally, for attention-based feature fusion, a Unified Feature Integrator (UFI) module is proposed. This module enhances quality perception capability by integrating global features and individual local quality features and introduces the Transformer to interact with the integrated quality features. Finally, feature regression is conducted to map the final features into the quality score. The performance of the proposed method is tested on four publicly available databases, and the experimental results show that the proposed method is superior compared with existing state-of-the-art no-reference PCQA methods in most cases.

As an important data form, the point cloud (PC) can present real-world objects in a 3D and realistic way. In order to measure the performance of acquisition, processing, and compression systems, it is crucial to propose a PC quality assessment (PCQA) method. Nowadays, a lot of full-reference PCQA methods have been developed, but it is obviously difficult to have wide applications. Usually, we should complete PCQA tasks without the original PC, so a no-reference PCQA method based on deep learning is proposed in this paper, which is consisting of adaptive feature extraction (AFE) module, local quality acquisition (LQA) module, and global quality acquisition (GQA) module. Specifically, MFE-Net is the core of the LQA module that combines a feature fusion mechanism with hierarchical feature extraction module and regression module, and makes full use of different hierarchical features. Besides, in the AFE module, hand-crafted features are extracted from local sampling clusters of distorted PCs by an adaptive approach, including the distance from the sampling center point, mean curvature, and gray level of the point, to replace the PC as the input of MFE-Net. Afterward, deep features can be learned based on such hand-crafted features to predict the local quality score, and the global quality can be obtained through aggregation in the GQA module. Experimental results on the two publicly available subjective datasets, SJTU-PCQA and CPCD2.0 show that the performance of the proposed method is superior to state-of-the-art PCQA methods.

The existing deep-learning based robust watermarking model generally applies a discriminator to form generative adversarial network (GAN) for increasing the quality of encoded images, and adopts a single encoder to embed watermark. However, GAN training is unstable, and the single encoder cannot fully adjust the watermarking distribution, thus affecting the watermarking performance. To address those limitations, this paper presents the multi-encoder based on conditional diffusion model (CDM) for robust image watermarking, namely, DiffW. To enhance the stability, the multi-encoder structure based on CDM replaces GAN for optimizing the watermarking distribution iteratively. Specifically, the operation of each timestep in the forward and reverse diffusion processes of the CDM is regarded as an encoder to overcome the shortcomings of the single encoder structure. At the training stage, under the guidance of the conditional noisy image, the forward process trains each encoder to fuse the image and watermark to generate high-quality encoded images. During the testing stage, only a small number of trained encoders of the forward process are used, so as to reduce the time complexity. Furthermore, to improve watermarking robustness, the channel attention module (CAM) is designed to extract main watermark features by mining channel correlations for multi-layer fusion, so that watermark can be embedded into imperceptible and texture areas. The experimental results reveal that compared with the existing watermarking model, the proposed DiffW can achieve better results in terms of watermarking invisibility and robustness.

As a crucial representation of 3D data, a point cloud (PC) can accurately capture the geometry, structure, and color information of objects. However, various quality problems arise owing to device noise, data acquisition errors, and compression algorithms, limiting the application of PCs. Therefore, assessing PC quality to determine its suitability for applications is a challenging task. In this work, a local and global structure-guided feature extraction and attention network (LGS-Net) is introduced for no-reference PC quality assessment (PCQA). This approach incorporates cluster construction (CC), local structure-guided cluster feature extraction (LSFE), and global structure-guided attention (GSA) modules. First, owing to the heightened sensitivity of the human visual system (HVS) to structural information, a graph filter is employed to identify high-frequency clusters. Within the LSFE module, a multiscale strategy is employed to ensure that structural information effectively influences both the geometry and color information. Simultaneously, the multiscale features within the cluster are dynamically fine-tuned using feature channel weight reassignment. To account for the impact of interclusters on overall quality, a GSA module is introduced to establish global dependencies between local clusters. This approach enables the extraction of final geometry, color, and structure information, which are ultimately used for accurate quality assessment. Extensive experimental results show that the proposed method outperforms the existing state-of-the-art PCQA methods using two publicly available subjective datasets.