Image steganography, a crucial technique for secure information transmission, faces the challenge of balancing embedding capacity with visual imperceptibility and security. Existing methods often struggle to maximize these metrics simultaneously, particularly when handling complex image details and achieving adaptive feature representation. To address this, we propose EctFormer, a novel deep steganography framework based on Image Hiding Empirical Mode Decomposition (IHEMD). EctFormer employs a compact autoencoder architecture with a key innovation: an integrated IHEMD module that adaptively decomposes images into physically meaningful intrinsic mode functions (IMFs) and residual components. This decomposition allows for superior feature representation and information embedding. Furthermore, we introduce an intrinsic mode loss function within a novel multi-image training strategy, achieving a remarkable embedding capacity of 96 bits per pixel. Experimental results on the DIV2K, COCO, and ImageNet datasets demonstrate EctFormer’s superior performance. Our method significantly improves PSNR (exceeding 17.00 dB for single-image tasks and 11.00 dB for multi-image tasks) while maintaining high SSIM values (above 0.99). These results surpass current state-of-the-art methods, validating the efficacy of our IHEMD-based approach and the proposed training strategy. EctFormer provides a new effective paradigm for image steganography and enables high-capacity, high-security covert communication. The code is available at https://github.com/lisen1129/EctFormer.

To enhance information security during transmission over public channels, images are frequently employed for binary data hiding. Nonetheless, data are vulnerable to distortion due to Joint Photographic Experts Group (JPEG) compression, leading to challenges in recovering the original binary data. Addressing this issue, this paper introduces a pioneering method for binary data hiding that leverages a combined spatial and channel attention Transformer, termed SCFformer, to withstand JPEG compression. This method employs a novel discrete cosine transform (DCT) quantization truncation mechanism during the hiding phase to bolster the stego image’s resistance to JPEG compression, using spatial and channel attention to conceal information in less perceptible areas, thereby enhancing the model’s resistance to steganalysis. In the extraction phase, the DCT quantization minimizes secret image loss during compression, facilitating easier information retrieval. The incorporation of scalable modules adds flexibility, allowing for variable-capacity data hiding. Experimental findings validate the high security, large capacity, and high flexibility of our scheme, alongside a marked improvement in binary data recovery post-JPEG compression, underscoring our method’s leading-edge performance.

Abstract Facial expression recognition (FER) is one of the popular research topics in computer vision. Most deep learning expression recognition methods perform well on a single dataset, but may struggle in cross‐domain FER applications when applied to different datasets. FER under cross‐dataset also suffers from difficulties such as feature distribution deviation and discriminator degradation. To address these issues, we propose a prototype‐oriented similarity transfer framework (POST) for cross‐domain FER. The bidirectional cross‐attention Swin Transformer (BCS Transformer) module is designed to aggregate local facial feature similarities across different domains, enabling the extraction of relevant cross‐domain features. The dual learnable category prototypes is designed to represent potential space samples for both source and target domains, ensuring enhanced domain alignment by leveraging both cross‐domain and specific domain features. We further introduce the self‐training resampling (STR) strategy to enhance similarity transfer. The experimental results with the RAF‐DB dataset as the source domain and the CK+, FER2013, JAFFE and SFEW 2.0 datasets as the target domains, show that our approach achieves much higher performance than the state‐of‐the‐art cross‐domain FER methods.

AMG is one of the most efficient and widely used methods for solving sparse linear systems. The computational process of AMG mainly consists of a series of iterative calculations of generalized sparse matrix-matrix multiplication (SpGEMM) and sparse matrix-vector multiplication (SpMV). Optimizing these sparse matrix calculations is crucial for accelerating solving linear systems. In this paper, we first focus on optimizing the SpGEMM algorithm in AmgX, a popular AMG library for GPUs. We propose a new algorithm called SpGEMM-upper, which achieves an average speedup of 2.02× on Tesla V100 and 1.96× on RTX 3090 against the original algorithm. Next, through experimental investigation, we conclude that no single SpGEMM library or algorithm performs optimally for most sparse matrices, and the same holds true for SpMV. Therefore, we build machine learning-based models to predict the optimal SpGEMM and SpMV used in the AMG calculation process. Finally, we integrate the prediction models, SpGEMM-upper, and other selected algorithms into a framework for adaptive sparse matrix computation in AMG. Our experimental results prove that the framework achieves promising performance improvements on the test set.