Accurately identifying correct correspondences (inliers) in two-view images is a fundamental task in computer vision. Recent studies usually adopt Graph Neural Networks or stack local graphs into global ones to establish neighborhood relations. However, the smoothing properties of Graph Convolutional Neural network (GCN) cause the model to fall into local extreme, which leads to the issue of indistinguishability between inliers and outliers. Especially when the initial correspondences contain a large number of incorrect correspondences (outliers), these studies suffer from severe performance degradation. To address the above issues and refocus perspective information on distinct features, we design a Consistency Guided ResFormer Network (CGR-Net) that uses consistent correspondences to guide model perspective focusing, thereby avoiding the negative impact of outliers. Specifically, we design an efficient Graph Score Calculation module, which aims to compute global graph scores by enhancing the representation of important features and comprehensively capturing the contextual relationships between correspondences. Then, we propose a Consistency Guided Correspondences Selection module to dynamically fuse global graph scores and consistency graphs and construct a novel consistency matrix to accurately recognize inliers. Extensive experiments on various challenging tasks demonstrate that our CGR-Net outperforms state-of-the-art methods. Our code is released at https://github.com/XiaojieLi11/CGR-Net.

We present a camera-based human body parameters measurement approach and develop a human postural assessment system. The approach combines the conventional contact measurement method and the non-contact measurement method to overcome some shortcomings in terms of time, expense, and professionalism in early methods. The entire measurement system consists of a computer, a high-definition camera, and the sticky points that are applied to the participant's body before the measurement. The camera captures the triple view image of human body. Then, the human body outline and the joint points of the human skeleton are extracted to locate the bone feature points. Finally, measurements and extractions of the human parameters are made. Experimental results demonstrate that the global postural assessment system provides quantitative guidance for human postural evaluation, and it completely changes how human postural is evaluated. The postural assessment system is significant for early diagnosis of diseases and medical rehabilitation treatment.

Recent years have seen the resurgence of self-supervised learning in visual representation thanks to Contrastive Learning and Masked Image Modeling. The existing self-supervised methods for skeleton-based action recognition typically learn feature invariance of the data only through contrastive learning. In this paper, we propose a contrast learning method combined with a temporal-masking mechanism of skeleton sequences to encourage the network able to learn action representations other than feature invariance, e.g., occlusion invariance, by implicitly reconstructing the masked sequences. However, the direct masking mechanism destroys the feature consistency of the samples, for which we propose Supervised Positive Sample Mining and self-attention module for embeddings to improve the generalization of the model. First of all, supervised contrastive learning can improve the robustness of models using prior knowledge of labels. Secondly, to avoid excessive masking mechanism that hinders the model from learning the correct occlusion invariance, a self-attention mechanism is necessary, which further discriminate the distance for each action class in the feature space. The results of various experimental protocols on NTU 60, NTU 120, PKU-MMD datasets demonstrate the advantages of our method and that our method outperforms the existing state-of-the-art contrastive methods. Code is available at https://github.com/ZZFCV/SASOiCLR .

As we know, image colorization is widely used in computer graphics and has become a research hotspot in the field of image processing. Current image colorization technology has the phenomenon of single coloring effect and unreal color, which is too complicated to be implemented and struggled to gain popularity. In this paper, a new method based on a convolution neural network is proposed to study the reasonable coloring of human images and ensures the realism of the coloring effect and the diversity of coloring at the same time. First, this paper selects about 5000 pictures of people and plants from the Imagenet dataset and makes a small dataset containing only people and backgrounds. Secondly, in order to obtain the image segmentation results, this paper improves the U-net network and carries out three times of down sampling and three times of up-sampling. Finally, we add the expanded convolution, and use the sigmoid activation function to replace the ReLU (The Rectified Linear Unit) activation function and put the BN (Batch Normalization) before the activation function. Experimental results show that our proposed image colorization algorithm based on the deep learning scheme can reduce the training time of the network and achieve higher quality segmentation results.

Soft tissue and organ modeling is the most critical function of any virtual surgical system. This study proposes a softness-based adaptive mesh refinement algorithm to simultaneously ensure realistic and real-time soft tissue simulation. The algorithm was constructed to consider that in a virtual surgery scenario, the surgical sites involve large deformation and thus require high simulation precision, whereas the nonsurgical sites involve small deformations and thus require low simulation precision. This study used the stomach lining as an example, applying mesh refinement in the deformation sites of the stomach lining to enhance the accuracy of the simulations. In addition, low mesh models were adopted for nonsurgical sites to ensure computing efficiency.

Virtual surface modeling has high scientific value and research significance as the most basic. According to the existing research, the operational speed is not enough, this paper constructs a three-dimensional(3D) reconstruction model of the cardiac surface with visualized human cardiac surface, using the point cloud after normal vector correction, to realize the rapid of the three-dimensional model of the cardiac surface.

Background: To simulate the anisotropic solid texture in 3D space, a kind of solid texture synthesis algorithm based on the samples that extends the 2D texture patch pasting technology to 3D hepatic artery solids is proposed. Methods: Firstly, the surface model is tetrahedral with Delaunay tetrahedral method. Secondly, the color attribute is used to generate the depth field and the depth field gradient direction is the first vector field, then the second vector field is defined by the arrow icon. The third vector is the product of the first two vectors. Finally, the vector field of the whole model is generated by Laplasse interpolation. Results & Conclusion: The experimental results showed that the proposed method can be used to synthesize the desired texture effect on any three-dimensional region, and it can simulate the internal structure of the hepatic artery and realize the real-time visualization of the cutting.

A complete 3D visualization method for virtual liver lesion model was proposed by taking patients’ abdomen CT slices as data source. Firstly, Gradient Vector Flow-Snake algorithm that combines with region force was adopted to fast and accurately extract the contour points in patients’ CT slices. Then 3D cloud data of these contour points was simplified uniformly, and a distance-field-based method of distribution field fitting for B-spline surface was presented to fast establish the lesion model roughly. Then, an interactive node fine tuning method Interactive Marching Nodes was proposed, so the model can be optimized according to requirements of users. Finally, realistic lesion model was generated through texture mapping. Experiments suggest that this approach is suitable for various kinds of lesions with the whole process taking only a few minutes, and the generated model has high precision, which is of some significance in the study of virtual surgeries.