The Vision Transformer (ViT) has achieved notable success in computer vision, with its variants widely validated across various downstream tasks, including semantic segmentation. However, as general-purpose visual encoders, ViT backbones often do not fully address the specific requirements of task decoders, highlighting opportunities for designing decoders optimized for efficient semantic segmentation. This paper proposes Strip Cross-Attention (SCASeg), an innovative decoder head specifically designed for semantic segmentation. Instead of relying on the conventional skip connections, we utilize lateral connections between encoder and decoder stages, leveraging encoder features as Queries in cross-attention modules. Additionally, we introduce a Cross-Layer Block (CLB) that integrates hierarchical feature maps from various encoder and decoder stages to form a unified representation for Keys and Values. The CLB also incorporates the local perceptual strengths of convolution, enabling SCASeg to capture both global and local context dependencies across multiple layers, thus enhancing feature interaction at different scales and improving overall efficiency. To further optimize computational efficiency, SCASeg compresses the channels of queries and keys into one dimension, creating strip-like patterns that reduce memory usage and increase inference speed compared to traditional vanilla cross-attention. Experiments show that SCASeg's adaptable decoder delivers competitive performance across various setups, outperforming leading segmentation architectures on benchmark datasets, including ADE20K, Cityscapes, COCO-Stuff 164k, and Pascal VOC2012, even under diverse computational constraints.

Medical image super-resolution (SR) is essential for enhancing diagnostic accuracy while reducing acquisition cost and scanning time. However, modeling both long-range anatomical structures and fine-grained frequency details with low computational overhead remains challenging. We propose FGMamba, a novel frequency-aware gated state-space model that unifies global dependency modeling and fine-detail enhancement into a lightweight architecture. Our method introduces two key innovations: a Gated Attention-enhanced State-Space Module (GASM) that integrates efficient state-space modeling with dualbranch spatial and channel attention, and a Pyramid Frequency Fusion Module (PFFM) that captures high-frequency details across multiple resolutions via FFT-guided fusion. Extensive evaluations across five medical imaging modalities (Ultrasound, OCT, MRI, CT, and Endoscopic) demonstrate that FGMamba achieves superior PSNR/SSIM while maintaining a compact parameter footprint (<0.75M), outperforming CNN-based and Transformerbased SOTAs. Our results validate the effectiveness of frequencyaware state-space modeling for scalable and accurate medical image enhancement. Source code and dataset will be made publicly available.

Privacy concerns have been persistently afflicting individuals within online social networks (OSNs), rendering privacy-preserving communications over the Internet with authentication especially important. Unfortunately, the guarantees of privacy and authenticity are not always provided in OSNs. Individuals are still facing the challenges of being deceived or exploited. To mitigate these issues, anonymous social networks (ASNs) have emerged as a remedy for OSNs, facilitating individuals to connect with others anonymously and authentically. Despite the existence of numerous and remarkable cryptographic primitives, there are no formal solutions for ASNs except for matchmaking encryption (ME), since ME can simultaneously provide various key functionalities, i.e. bilateral access control, identity anonymity, and message authentication, to address the requirements of ASNs. In this paper, we design a system for ASNs by adopting fuzzy identity-based matchmaking encryption (fuzzy IB-ME), and the proposed scheme in this work is highly efficient. The scheme also realizes adaptive security in generic group model (GGM), which is generally adopted in pairing-based cryptography. The proposed ASNs system offers various advantages compared to the previous solutions, including 1) bilateral access control, 2) enhanced security, 3) high scalability, and 4) high efficiency. In addition to theoretical evaluations, we conduct extensive experiments to evaluate our scheme’s computational and storage efficiency. These evaluations indicate that our solution outperforms previous solutions and as well as preserves many desired functionalities.

Images captured in challenging environments-such as nighttime, smoke, rainy weather, and underwater-often suffer from significant degradation, resulting in a substantial loss of visual quality. The effective restoration of these degraded images is critical for the subsequent vision tasks. While many existing approaches have successfully incorporated specific priors for individual tasks, these tailored solutions limit their applicability to other degradations. In this work, we propose a universal network architecture, dubbed "ReviveDiff", which can address various degradations and restore images to their original quality by enhancing and restoring their details. Our approach is inspired by the observation that, unlike degradation caused by movement or electronic issues, quality degradation under adverse conditions primarily stems from natural media (such as fog, water, and low luminance), which generally preserves the original structures of objects. To restore the quality of such images, we leveraged the latest advancements in diffusion models and developed ReviveDiff to restore image quality from both macro and micro levels across some key factors determining image quality, such as sharpness, distortion, noise level, dynamic range, and color accuracy. We rigorously evaluated ReviveDiff on seven benchmark datasets covering five types of degrading conditions: Rainy, Underwater, Low-light, Smoke, and Nighttime Hazy. Our experimental results demonstrate that ReviveDiff outperforms the state-of-the-art methods both quantitatively and visually.

The Vision Transformer (ViT) has achieved remarkable success in computer vision due to its powerful token mixer, which effectively captures global dependencies among all tokens. However, the quadratic complexity of standard self-attention with respect to the number of tokens severely hampers its computational efficiency in practical deployment. Although recent hybrid approaches have sought to combine the strengths of convolutions and self-attention to improve the performance-efficiency trade-off, the costly pairwise token interactions and heavy matrix operations in conventional self-attention remain a critical bottleneck. To overcome this limitation, we introduce S2AFormer, an efficient Vision Transformer architecture built around a novel Strip Self-Attention (SSA) mechanism. Our design incorporates lightweight yet effective Hybrid Perception Blocks (HPBs) that seamlessly fuse the local inductive biases of CNNs with the global modeling capability of Transformer-style attention. The core innovation of SSA lies in simultaneously reducing the spatial resolution of the key ( $K$ ) and value ( $V$ ) tensors while compressing the channel dimension of the query ( $Q$ ) and key ( $K$ ) tensors. This joint spatial-and-channel compression dramatically lowers computational cost without sacrificing representational power, achieving an excellent balance between accuracy and efficiency. We extensively evaluate S2AFormer on a wide range of vision tasks, including image classification (ImageNet-1K), semantic segmentation (ADE20K), and object detection/instance segmentation (COCO). Experimental results consistently show that S2AFormer delivers substantial accuracy improvements together with superior inference speed and throughput across both GPU and non-GPU platforms, establishing it as a highly competitive solution in the landscape of efficient Vision Transformers.

Low-light environments can cause a variety of complex degradation problems, which result in poor visibility in images. As a classical vision task, low-light image enhancement has attracted an increasing interest in the research community. However, the existing methods tend to require a large number of parameters, making them difficult to implement and optimize, especially on resource-constrained devices. In this paper, we mainly focus on the lightweight of the method and propose a novel end-to-end two-stage CNN-ViT architecture (HSINet) to learn hierarchical semantic information (HSI) from low-light images efficiently. The HSINet consists of two stages: the first stage is a CNN-based low-level semantic (LS) Stage, and the second stage is ViT-based high-level semantic (HS) Stage. The LS Stage contains an efficient multi-scale convolution block, MLS Block, for low-level semantic information extraction. The HS stage, on the other hand, aims to learn the high-level semantic features via ViT's excellent global-learning capability. We propose a hierarchical Swin Transformer-based block, HS Block, to gradually enlarge Swin Transformer's window size as the network becomes deeper, to learn hierarchical high-level semantic information. Benefiting from the efficient architecture, our model only contains 0.6M parameters, far fewer than the existing SOTAs. We evaluated the method on three challenging benchmark datasets: LOL, VE-LOL, and MIT-Adobe FiveK, using three popular evaluation metrics. The quantitative and qualitative results both show that the proposed method not only outperforms the state of the arts in terms of PSNR, SSIM, LPIPS, and visual effects, but also with better efficiency.

The existing saliency detection models based on RGB colors only leverage appearance cues to detect salient objects. Depth information also plays a very important role in visual saliency detection and can supply complementary cues for saliency detection. Although many RGB-D saliency models have been proposed, they require to acquire depth data, which is expensive and not easy to get. In this paper, we propose to estimate depth information from monocular RGB images and leverage the intermediate depth features to enhance the saliency detection performance in a deep neural network framework. Specifically, we first use an encoder network to extract common features from each RGB image and then build two decoder networks for depth estimation and saliency detection, respectively. The depth decoder features can be fused with the RGB saliency features to enhance their capability. Furthermore, we also propose a novel dense multiscale fusion model to densely fuse multiscale depth and RGB features based on the dense ASPP model. A new global context branch is also added to boost the multiscale features. Experimental results demonstrate that the added depth cues and the proposed fusion model can both improve the saliency detection performance. Finally, our model not only outperforms state-of-the-art RGB saliency models, but also achieves comparable results compared with state-of-the-art RGB-D saliency models.