Efficient image super-resolution (SR) models are essential for achieving high-quality image reconstruction with reduced computational complexity, particularly in resource-constrained environments. In this paper, we introduce a novel self-attention mechanism, Broadcast-Gated Attention with Identity Adaptive Integration (BGAI). Then, based on this mechanism, we design a lightweight super-resolution network that achieves state-of-the-art performance with minimal computational cost. By observing the sparsity and convergence properties of self-attention, BGAI optimizes computational resource utilization through the effective broadcasting of meaningful features across attention heads and network layers. A key innovation in BGAI is the Broadcast-Gated Multi-head Self-Attention (BGMSA) mechanism, which employs a dedicated head to capture and integrate long-range dependencies, broadcasting this broader contextual information to local attention heads. This design enhances long-range interaction modeling while minimizing redundant computations. Additionally, the Identity Attention Adaptive Integration (IAAI) mechanism facilitates efficient feature propagation by leveraging the continuity in dependencies across layers, with a focus on dynamic variations to improve representational efficiency and accelerate convergence. Comprehensive experiments on standard benchmarks demonstrate that BGAI achieves high-fidelity super-resolution while reducing the number of parameters and FLOPs by up to 35% compared with existing lightweight methods. These results establish BGAI as a robust and scalable solution for resource-efficient SR, with significant potential for deployment in real-world, high-resolution image processing applications. The code and trained models are publicly available at https://github.com/bbbolt/BGAI.

Recent advancements in subject-driven image generation have led to zero-shot generation, yet precise selection and focus on crucial subject representations remain challenging. Addressing this, we introduce the SSR-Encoder, a novel architecture designed for selectively capturing any subject from single or multiple reference images. It responds to various query modalities including text and masks, without necessitating test-time fine-tuning. The SSR-Encoder combines a Token-to-Patch Aligner that aligns query inputs with image patches and a Detail-Preserving Subject Encoder for extracting and preserving fine features of the subjects, thereby generating subject embeddings. These embeddings, used in conjunction with original text embeddings, condition the generation process. Characterized by its model generalizability and efficiency, the SSR-Encoder adapts to a range of custom models and control modules. Enhanced by the Embedding Consistency Regularization Loss for improved training, our extensive experiments demonstrate its effectiveness in versatile and high-quality image generation, indicating its broad applicability. Project page: ssr-encoder.github.io

Currently, personalized image generation methods mostly require considerable time to finetune and often overfit the concept resulting in generated images that are similar to custom concepts but difficult to edit by prompts. We propose an effective and fast approach that could balance the text-image consistency and identity consistency of the generated image and reference image. Our method can generate personalized images without any fine-tuning while maintaining the inherent text-to-image generation ability of diffusion models. Given a prompt and a reference image, we merge the custom concept into generated images by manipulating cross-attention and self-attention layers of the original diffusion model to generate personalized images that match the text description. Comprehensive experiments highlight the superiority of our method.

Hyperspectral (HS) and multispectral (MS) images’ fusion aims to improve their spatial resolutions and circumvent the main limitation of HS sensors. However, existing HS–MS fusion methods account for spectral variability fail to consider the global spectral correlation. To overcome this problem, this letter presents a novel joint weighted Schatten-p norm and spatial smoothness regularization for HS–MS fusion account for both spatial and spectral changes. First, the relationship between the spectral variability and the spectral signatures is formulated as an explicit parametric model. Second, to preserve the inherent correlation among the bands, we design a weighted Schatten-p ($ 0\lt p\lt 1 $) norm regularization method, which considers the importance of different components. Third, a spatial smoothness regularization term is exploited to reconstruct the spatial details. Finally, an iterative procedure based on the framework of alternating direction method of multipliers (ADMM) is designed to solve the resulting optimization problem. Extensive experiments on both synthetic and real datasets demonstrate that the proposed method outperforms six state-of-the-art methods from visual and quantitative assessments. The datasets and results are released inhttp://github.com/phan1007/WSGS.

Although the fast tool servo (FTS) exhibits unique advantages for surface texturing, its intrinsic form accuracy is capped by the position and motion accuracy of the manufacturing system. In this article, a novel FTS combining the sensing and control capability of 2-DOF tool position and cutting force is developed to enable the high-performance adaptive surface texturing. The proposed design of FTS features a motion-decoupled symmetric configuration and functions via leveraging the system dynamics and online position monitoring of two pairs of actuation-force sensing observers. Analytical models and the differential evolution algorithm are established to describe the essential working performances and guide the optimal structural parameter design, which are validated using finite element simulation. The performance assessments suggest positioning resolution around 20 nm and force-sensing resolution at the millinewton level for both cutting and thrust directions, along with motion strokes of tens of micrometers and resonant frequency of up to 4172 Hz. Moreover, the adaptive texturing of microgrooved surfaces with consistent geometry is demonstrated without prior knowledge of workpiece shapes. The outcomes of this article contribute to developing the next-generation of intelligent FTS for flexible and intelligent manufacturing.

Strong tracking Kalman filter is proposed to address the performance degradation and divergence caused by process model uncertainty. However, strong tracking filters depend on prior information about the measurement noise, which is frequently unknown or time-varying in real-world applications. This paper proposed an adaptive strong tracking cubature Kalman filtering algorithm with unknown measurement noise covariance. First, we introduce a modified pseudo-measurement-based covariance estimation method. It estimates measurement covariance by calculating the second-order mutual difference between the measurement sequence and pseudo measurement sequence. Second, we propose a filter divergence detecting method to help decide when to adjust the prediction error covariance matrix. The accuracy of the measurement noise covariance matrix estimating method and the effectiveness of filter divergence detecting methods have been proved by simulation results, respectively. As a result, the proposed filter outperforms several other algorithms in precision or robustness with inaccurate measurement noise covariance.

Text-to-face (T2F) generation is an emerging research hot spot in multimedia, and its main challenge lies in the high fidelity requirement of generated portraits. Many existing works resort to exploring the latent space in a pre-trained generator, e.g., StyleGAN, which has obvious shortcomings in efficiency and generalization ability. In this paper, we propose a generative network for open-ended text-to-face generation, which is termed OpenFaceGAN. Differing from existing StyleGAN-based methods, OpenFaceGAN constructs an effective multi-modal latent space that directly converts the natural language description into a face. This mapping paradigm can fit the real data distribution well and make the model capable of open-ended and even zero-shot T2F generation. Our method improves the inference speed by an order of magnitude, e.g., 294 times than TediGAN. Based on OpenFaceGAN, we further explore text-guided face manipulation (editing). In particular, we propose a parameterized module, OpenEditor, to automatically disentangle the target latent code and update the original style information. OpenEditor also makes OpenFaceGAN directly applicable for most manipulation instructions without example-dependent searches or optimizations, greatly improving the efficiency of face manipulation. We conduct extensive experiments on two benchmark datasets namely Multi-Modal CelebA-HQ and Face2Text-v1.0. The experimental results not only show the superior performance of OpenFaceGAN to the existing T2F methods in both image quality and image-text matching degree but also greatly confirm its outstanding ability in the zero-shot generation. Codes will be released at: \textcolormagenta \urlhttps://github.com/pengjunn/OpenFace

Different image fusion systems have been developed to deal with the massive amounts of image data for different applications, such as remote sensing, computer vision, and environment monitoring. However, the generalizability and versatility of these fusion systems remain unknown. This article proposes an efficient regularization framework to achieve different kinds of fusion tasks accounting for the spatiospectral and spatiotemporal variabilities of the fusion process. A joint minimization functional is developed by taking an advantage of a composite regularizer for enforcing joint sparsity in the gradient domain and the frame domain. The proposed composite regularizer is composed of the Hessian Schatten-norm regularization and contourlet-based regularization terms. The resulting problems are solved by the alternating direction method of multipliers (ADMM). The effectiveness of the proposed method is validated in a variety of image fusion experiments: 1) hyperspectral (HS) and panchromatic image fusion; 2) HS and multispectral image fusion; 3) multitemporal image fusion (MIF); and 4) multi-image deblurring. Results show promising performance compared with state-of-the-art fusion methods.

A robust adaptive filter is proposed by using the variational Bayesian (VB) inference to extended target tracking with heavy‐tailed noise in clutter. An explicit distribution is used to describe the non‐Gaussian heavy‐tailed noise based on Student's t ‐distribution. The need for arbitrary decisions is then eliminated, and the robust operation is provided which is less sensitive to extreme observation. Moreover, an approximate measurement update using the analytical techniques of VB methods is derived to approximate the posterior states at each time step. To obtain a more accurate result, clutter estimation is also integrated considering the uncertainty of target tracking in a cluttered environment. The performance of the proposed algorithm is demonstrated with simulated data.

Image restoration is a difficult and challenging problem in various imaging applications. However, despite of the benefits of a single overcomplete dictionary, there are still several challenges for capturing the geometric structure of image of interest. To more accurately represent the local structures of the underlying signals, we propose a new problem formulation for sparse representation with block-orthogonal constraint. There are three contributions. First, a framework for discriminative structured dictionary learning is proposed, which leads to a smooth manifold structure and quotient search spaces. Second, an alternating minimization scheme is proposed after taking both the cost function and the constraints into account. This is achieved by iteratively alternating between updating the block structure of the dictionary defined on Grassmann manifold and sparsifying the dictionary atoms automatically. Third, Riemannian conjugate gradient is considered to track local subspaces efficiently with a convergence guarantee. Extensive experiments on various datasets demonstrate that the proposed method outperforms the state-of-the-art methods on the removal of mixed Gaussian-impulse noise.

This study proposes a recursive solution to target joint detection, tracking, and classification (DTC) based on finite set statistics (FISST) and generalised Bayesian risk. A new Bayesian risk is defined involving the costs of target existence probability estimation (detection), state estimation (tracking), and classification. The estimates and costs are calculated within the FISST framework for different hypotheses and decisions of the target class, and the optimal solution is then derived to minimise the new Bayesian risk. As different costs are unified, the inter‐dependence of DTC is considered, and these three sub‐problems are solved jointly. The effects of the parameters of the new Bayesian risk are also analysed. Simulations show that the authors’ method has a better overall performance compared with traditional methods.