Numerous CNN-based algorithms have been proposed to reconstruct high-quality face images. However, the inability of convolution operation to model long-distance relationships limits the performance of the CNN-based methods. Moreover, in the high-resolution (HR) image reconstruction stage, with the well decoded feature representations, more efficient architecture design can be explored to synthesize pixel-level image details. In this work, we propose a spatial attention-guided CNN-Transformer aggregation network (SCTANet) for face image super-resolution (FSR) tasks. The core component in the deep feature extraction stage is the Hybrid Attention Aggregation (HAA) block. The HAA block has two parallel paths, one for the Residual Spatial Attention (RSA) block, the other for the Multi-scale Patch embedding and Spatial-attention Masked Transformer (MPSMT) block. The HAA block combines the strengths of CNN and transformer to effectively exploit both local and global information. For the reconstruction stage, we propose to use the Sub-pixel MLP-based Upsampling (SMU) module instead of the conventional CNN architecture. The SMU module promotes the reconstruction of pixel-level image details and reduces computational complexity. Extensive experiments on both synthetic and real-world face datasets demonstrate the superiority of our proposed SCTANet over state-of-the-art methods.

Existing shadow removal methods often struggle with two problems: color inconsistencies in shadow areas and artifacts along shadow boundaries. To address these two problems, we propose a novel shadow mask-based semantic-aware network (S ² Net) that uses shadow masks as guidance for shadow removal. The color inconsistency problem is solved in two steps. First, we use a series of semantic-guided dilated residual (SDR) blocks to transfer statistical information from non-shadow areas to shadow areas. The shadow mask-based semantic transformation (SST) operation in SDR enables the network to remove shadows while keeping non-shadow areas intact. Then, we design a refinement block by incorporating semantic knowledge of shadow masks and applying the learned modulated convolution kernels to get traceless and consistent output. To remove artifacts along shadow boundaries, we propose a newly designed boundary loss. The boundary loss encourages spatial coherence around shadow boundaries. By including the boundary loss as part of the loss function, a significant portion of artifacts along shadow boundaries can be removed. Extensive experiments on the ISTD, ISTD+, SRD and SBU datasets show our S ² Net outperforms existing shadow removal methods.

Compared to wide-field telescopes, small-field detection systems have higher spatial resolution, resulting in stronger detection capabilities and higher positioning accuracy. When detecting by small fields in synchronous orbit, both space debris and fixed stars are imaged as point targets, making it difficult to distinguish them. In addition, with the improvement in detection capabilities, the number of stars in the background rapidly increases, which puts higher requirements on recognition algorithms. Therefore, star detection is indispensable for identifying and locating space debris in complex backgrounds. To address these difficulties, this paper proposes a real-time star extraction method based on adaptive filtering and multi-frame projection. We use bad point repair and background suppression algorithms to preprocess star images. Afterwards, we analyze and enhance the target signal-to-noise ratio (SNR). Then, we use multi-frame projection to fuse information. Subsequently, adaptive filtering, adaptive morphology, and adaptive median filtering algorithms are proposed to detect trajectories. Finally, the projection is released to locate the target. Our recognition algorithm has been verified by real star images, and the images were captured using small-field telescopes. The experimental results demonstrate the effectiveness of the algorithm proposed in this paper. We successfully extracted hip-27066 star, which has a magnitude of about 12 and an SNR of about 1.5. Compared with existing methods, our algorithm has advantages in both recognition rate and false-alarm rate, and can be used as a real-time target recognition algorithm for space-based synchronous orbit detection payloads.

Compression artifacts removal methods based on convolutional neural networks have attracted great attention. However, most existing methods require a specific trained model for a specific compression quality factor (QF), which inevitably leads to resource-consuming. Unfortunately, the QF is unknown in most practical applications, so it is intractable to choose a suitable model. In this work, we experimentally analyze the relationship between compression index estimation and compression artifacts removal. Based on the connection between them, we couple compression index estimation with compression artifacts removal into a unified network. A network named CRESNet is proposed, working for a wide range of QFs by integrating channel regulation with an exit strategy. Specifically, CRESNet adopts a multi-stage progressive structure with an exit strategy embedded to automatically select the optimal exit stage according to the estimated compression index reflecting the difficulty of the input sample. Benefiting from the exit strategy, CRESNet removes artifacts from slightly compressed images through a simple process while doing an elaborate process for severely compressed images. Furthermore, a compression-information-guided channel regulation (CICR) mechanism is developed to adaptively regulate feature maps based on the estimated compression index. CRESNet achieves a more elegant trade-off between artifacts removal and detail preservation in a resource-efficient manner. Experiments demonstrate that CRESNet achieves state-of-the-art performance.

Reliable pose estimation for non-cooperative spacecraft is a key technology for in-orbit service and active debris removal missions. Utilizing deep learning techniques for processing monocular camera images is effective and is a hotspot of current research. To reduce errors and improve model generalization, researchers often design multi-head loss functions or use generative models to achieve complex data augmentation, which makes the task complex and time-consuming. We propose a pyramid vision transformer spatial-aware keypoints regression network and a stereo-aware augmentation strategy to achieve robust prediction. Specifically, we primarily use the eight vertices of a cuboid satellite body as landmarks and the observable surfaces can be transformed by, respectively, using the pose labels. The experimental results on the SPEED+ dataset show that by using the existing EPNP algorithm and pseudo-label self-training method, we can achieve high-precision pose estimation for target cross-domains. Compared to other existing methods, our model and strategy are more straightforward. The entire process does not require the generation of new images, which significantly reduces the storage requirements and time costs. Combined with a Kalman filter, the robust and continuous output of the target position and attitude is verified by the SHIRT dataset. This work realizes deployment on mobile devices and provides strong technical support for the application of an automatic visual navigation system in orbit.

Convolutional neural network-based image super-resolution methods have achieved great success in recent years. However, the huge memory and computational costs make most of the existing methods difficult to be applied to resource-constrained scenarios such as edge devices. To tackle this problem, we propose a generic, lightweight and efficient feature fusion block to replace the commonly used 1*1 convolution. In addition, we propose the enhanced shallow residual blocks to improve the super-resolution performance. By combining these two novel blocks, we design an efficient similar feature fusion network for single image super-resolution, based on the observation that cross-layer features of the same channel usually have high similarities. The similar feature fusion block utilizes similarity as a guide for feature clustering, enabling efficient and high-performance cross-layer feature fusion. On the other hand, the enhanced shallow residual blocks are used as the base feature extraction model for the network to improve super-resolution performance in conjunction with the feature fusion module. In the enhanced shallow residual blocks, we combine convolution with identity connection to maintain the similarity of cross-layer features that are fed into the similar feature fusion block. The spatial attention mechanism is also introduced to reinforce the useful spatial features. Experimental results on the benchmark datasets show that the proposed method can achieve comparable results to state-of-the-art methods with a small number of parameters.