To improve the accuracy of color image completion with missing entries, we present a recovery method based on generalized higher-order scalars. We extend the traditional second-order matrix model to a more comprehensive higher-order matrix equivalent, called the “t-matrix” model, which incorporates a pixel neighborhood expansion strategy to characterize the local pixel constraints. This “t-matrix” model is then used to extend some commonly used matrix and tensor completion algorithms to their higher-order versions. We perform extensive experiments on various algorithms using simulated data and publicly available images. The results show that our generalized matrix completion model and the corresponding algorithm compare favorably with their lower-order tensor and conventional matrix counterparts.

Facial expression recognition (FER) in the wild is challenging due to various unconstrained conditions, i.e., occlusions and head pose variations. Previous methods tend to improve the performance of facial expression recognition through resorting to holistic methods or coarse local-based methods, while ignoring the local fine-grained feature structure knowledge and the correlation between features. In this paper, we propose a Fine-Grained Association Graph Representation (FG-AGR) framework which can capture the local fine-grained facial expression representation. Firstly, an Adaptive Salient Region Induction (ASRI) is designed for adaptively highlighting the local saliency regions of facial expressions combined with spatial location information. Based on this, a Local Fine-grained Feature Extraction (LFFE) based on Visual Transformers is introduced to further extract fine but discriminative fine-grained features of saliency regions. Thirdly, an Adaptive Graph Association Reasoning (AGAR) based on Graph Convolutional Network is constructed to learn associated fine-grained feature combinations. Extensive experiments demonstrate that our FG-AGR achieves superior performance compared to the state-of-the-art methods with 90.81% on RAF-DB, 64.91% on AffectNet-7, 60.69% on AffectNet-8 and 91.09% on FERPlus.

In contrast to defect-free fabric images with macro-homogeneous textures and regular patterns, fabric images with defect are characterized by defect regions that are salient and sparse among the redundant background. Therefore, as an effective tool for separating an image into redundant part (the background) and sparse part (the defect), low-rank decomposition model provides an ideal solution for pattered fabric defect detection. In this paper, a novel patterned method for fabric defect detection is proposed based on a novel texture descriptor and low-rank decomposition model. Firstly, an efficient second-order orientation-aware descriptor, denoted as GHOG, is designed by combining Gabor and Histogram of Oriented Gradient (HOG). In addition, a spatial pooling strategy based on human vision mechanism is utilized to further improve the discrimination ability of the proposed descriptor. The proposed texture descriptor can make the defect-free image blocks lay in a low-rank subspace, while the defective image blocks are deviated from this subspace. Then a constructed low-rank decomposition model divides the feature matrix generated from all the image blocks into low-rank part which represents defect-free background and sparse part which represents sparse defects. In addition, a non-convex log det as a smooth surrogate function is utilized to improve the efficiency of the constructed low-rank model. Finally, the defects are localized by segmenting the saliency map generated by the sparse matrix. Qualitative results and quantitative evaluations results demonstrate that the proposed method improves the detection accuracy and self-adaptivity comparing with the state-of-the-art methods.

Fabric defect detection plays a key role in the quality control of textiles. Existing fabric defect detection methods adopt traditional pattern recognition methods; however, these methods lack adaptability and present poor detection performance. Because biological vision system has the ability to quickly locate salient objects, we propose a novel fabric defect detection algorithm based on biological vision modeling by simulating the mechanism of biological visual perception. First, a distinct, efficient, and robust feature descriptor from the biological modeling of P ganglion cells, which was proposed in our previous work, is adopted to improve the representation of fabric images with complex textures. To account for the low-rank and sparsity characteristics of biological vision, the low-rank representation (LRR) technique is adopted to model biological visual saliency, and it can decompose the fabric image into backgrounds and salient defect objects. Meanwhile, dictionary learning and Laplacian regularization are integrated into the LRR model as follows: 1) dictionary learning is used to denoise the saliency map; and 2) Laplacian regularization enlarges the gaps between defective regions and the background. Finally, the linearized accelerated direction method with adaptive penalty is adopted to solve the proposed model. The experimental results emphasize that the proposed algorithm has good detection performance for plain or twill fabrics with simple textures as well as for patterned fabrics with complex textures. Moreover, the proposed method is superior to the state-of-the-art methods in terms of its adaptability and detection efficiency.

In this paper, a novel mechanism is studied to improve the performance of the channel state information (CSI) feedback in massive multiple-input multiple-output (MIMO) systems. The proposed mechanism encompasses convolutional neural network (CNN)-based CSI compression and reconstruction structure. In this structure, the long-short term memory (LSTM) is adopted to learn temporal correlation of channels, and then, an attention mechanism is developed to perceive local information and automatically weight feature information. In addition, the CNN framework is further adjusted to reduce the number of training parameters and accelerate CSI recovery. The CNN structure with optimal training parameters can be achieved via offline iterative training and learning based on various training datasets. Comparative experimental studies demonstrate the effectiveness of the proposed approach that the trained CNN can obtain the higher feedback accuracy and better system performance in massive MIMO CSI online feedback reconstruction. Moreover, the proposed scheme in the less parameters-based neural network owns a higher performance with lower computational complexity compared to the conventional algorithms.