Heterogeneous Graph Neural Networks (HGNNs), have demonstrated excellent capabilities in processing heterogeneous information networks. Self-supervised learning on heterogeneous graphs, especially contrastive self-supervised strategy, shows great potential when there are no labels. However, this approach requires the use of carefully designed graph augmentation strategies and the selection of positive and negative samples. Determining the exact level of similarity between sample pairs is non-trivial.To solve this problem, we propose a novel self-supervised Heterogeneous graph neural network with Optimal Transport (HGOT) method which is designed to facilitate self-supervised learning for heterogeneous graphs without graph augmentation strategies. Different from traditional contrastive self-supervised learning, HGOT employs the optimal transport mechanism to relieve the laborious sampling process of positive and negative samples. Specifically, we design an aggregating view (central view) to integrate the semantic information contained in the views represented by different meta-paths (branch views). Then, we introduce an optimal transport plan to identify the transport relationship between the semantics contained in the branch view and the central view. This allows the optimal transport plan between graphs to align with the representations, forcing the encoder to learn node representations that are more similar to the graph space and of higher quality. Extensive experiments on four real-world datasets demonstrate that our proposed HGOT model can achieve state-of-the-art performance on various downstream tasks. In particular, in the node classification task, HGOT achieves an average of more than 6\% improvement in accuracy compared with state-of-the-art methods.

ABSTRACT Facial expression plays a crucial role during interactions with people. Previous studies on facial expression recognition (FER) have mainly focused on adults, while there are few studies on FER for infants. Due to the apparent differences in facial proportions and facial contours between infants and adults, the FER studies for infants could not be conducted on existing expression datasets. In order to study infant facial expressions in‐depth, we create the infant facial expression recognition (IFER) dataset by collecting 10,240 infant images. Since infants' faces have smooth facial lines and weak sharpness, the inter‐class similarity of facial expressions is higher than adults, and the existing networks for facial expression recognition lack attention to inter‐class similarity. To address the above problems, we propose an effective infant facial expression recognition network named EFNet. In the first stage, the convolutional neural network (CNN) branch and the self‐attention branch extract the overall features of infants' faces. In the second stage, we propose the self‐adaptive attentional centre loss (SACL). The SACL uses the extracted feature maps as contexts to estimate the weights by an attention mechanism and then applies the attentional weights to guide the centre loss. Overall, the SACL facilitates inter‐class separateness and intra‐class compressiveness of related information in an embedding space. The state‐of‐the‐art results on the IFER dataset confirm the remarkable effectiveness of the EFNet.

Fundus disease is a complex and universal disease involving a variety of pathologies. Its early diagnosis using fundus images can effectively prevent further diseases and provide targeted treatment plans for patients. Recent deep learning models for classification of this disease are gradually emerging as a critical research field, which is attracting widespread attention. However, in practice, most of the existing methods only focus on local visual cues of a single image, and ignore the underlying explicit interaction similarity between subjects and correlation information among pathologies in fundus diseases. In this paper, we propose a novel label-aware dual graph neural networks for multi-label fundus image classification that consists of population-based graph representation learning and pathology-based graph representation learning modules. Specifically, we first construct a population-based graph by integrating image features and non-image information to learn patient's representations by incorporating associations between subjects. Then, we represent pathologies as a sparse graph where its nodes are associated with pathology-based feature vectors and the edges correspond to probability of the co-occurrence of labels to generate a set of classifier scores by the propagation of multi-layer graph information. Finally, our model can adaptively recalibrate multi-label outputs. Detailed experiments and analysis of our results show the effectiveness of our method compared with state-of-the-art multi-label fundus image classification methods.

Graph neural networks (GNNs) have demonstrated remarkable success for semisupervised node classification. However, these GNNs are still limited to the conventionally semisupervised framework and cannot fully leverage the potential value of large numbers of unlabeled samples. The pseudolabeling method in semisupervised learning (SSL) is widely recognized because it can clearly leverage unlabeled samples. Nevertheless, the existing pseudolabeling methods usually utilize a fixed threshold for all classes and only use a portion of unlabeled samples (ones with high prediction confidence), which leads to class imbalance and low data utilization. To solve these problems, we propose GNNs with adaptive confidence discrimination (ACDGNN) to fully utilize unlabeled samples for facilitating semisupervised node classification. Specifically, an adaptive confidence discrimination module is designed to divide all unlabeled nodes into two subsets by comparing their confidence scores with the adaptive confidence threshold at each training epoch. Then, different constraint strategies for two subset nodes are employed. Unlabeled nodes with high confidence are used to iteratively expand the label set, while ones with low confidence learn discriminative features by applying contrastive learning. Validated by extensive experiments, the proposed ACDGNN delivers significant accuracy gains over the previous SOTAs: an average improvement of 2.0% on all datasets and 5.7% on the Flickr dataset in particular.

Graph contrastive learning (GCL), as a typical self-supervised learning paradigm, has been able to achieve promising performance without labels and gradually attracts much attention. Graph-level method aims to learn representations of each graph by contrasting two augmented graphs. Previous studies usually simply apply contrastive learning to keep the embeddings of augmented views from the same anchor graph (positive pairs) close to each other, as well as separate the embeddings of augmented views from different anchor graphs (negative pairs). However, it is well-known that the structure of graph is always complex and multiscale, which gives rise to a fundamental question: after graph augmentation, will the previous assumption still hold in reality? Through experimental analytics, we find that the semantic information of two augmented graphs from the same anchor graph may be not consistent, and whether two augmented graphs are positive or negative sample pairs is highly correlated with the multiscale structure of the graph. Based on this observation, we then propose a multiscale subgraph contrastive learning method, named MSSGCL, which can characterize the fine-grained semantic information. Specifically, we generate global and local views at different scales based on subgraph sampling and construct multiple contrastive relationships according to their semantic associations to provide richer self-supervised information. Furthermore, to further improve the generalization performance of the model, we propose an extended model called MSSGCL++. It adopts an asymmetric structure to avoid pushing semantically similar negative samples far away. We further introduce adversarial training to perturb the augmented view and thus construct a more difficult self-supervised training task. Finally, a min-max saddle point problem is optimized and the "free" strategy is used to speed up the training process. Extensive experiments and parametric analysis on 16 real-world graph classification datasets confirm the effectiveness of our proposed approach. Compared with state of the art (SOTA) method, our method achieves improvements of 2% and 1.6% in unsupervised and transfer learning settings, respectively.

Heterogeneous graphs with multiple types of nodes and link relationships are ubiquitous in many real-world applications. Heterogeneous graph neural networks (HGNNs) as an efficient technique have shown superior capacity of dealing with heterogeneous graphs. Existing HGNNs usually define multiple meta-paths in a heterogeneous graph to capture the composite relations and guide neighbor selection. However, these models only consider the simple relationships (i.e., concatenation or linear superposition) between different meta-paths, ignoring more general or complex relationships. In this article, we propose a novel unsupervised framework termed Heterogeneous Graph neural network with bidirectional encoding representation (HGBER) to learn comprehensive node representations. Specifically, the contrastive forward encoding is firstly performed to extract node representations on a set of meta-specific graphs corresponding to meta-paths. We then introduce the reversed encoding for the degradation process from the final node representations to each single meta-specific node representations. Moreover, to learn structure-preserving node representations, we further utilize a self-training module to discover the optimal node distribution through iterative optimization. Extensive experiments on five open public datasets show that the proposed HGBER model outperforms the state-of-the-art HGNNs baselines by 0.8%-8.4% in terms of accuracy on most datasets in various downstream tasks.

Graph-level contrastive learning, aiming to learn the representations for each graph by contrasting two augmented graphs, has attracted considerable attention. Previous studies usually simply assume that a graph and its augmented graph as a positive pair, otherwise as a negative pair. However, it is well known that graph structure is always complex and multi-scale, which gives rise to a fundamental question: after graph augmentation, will the previous assumption still hold in reality? By an experimental analysis, we discover the semantic information of an augmented graph structure may be not consistent as original graph structure, and whether two augmented graphs are positive or negative pairs is highly related with the multi-scale structures. Based on this finding, we propose a multi-scale subgraph contrastive learning architecture which is able to characterize the fine-grained semantic information. Specifically, we generate global and local views at different scales based on subgraph sampling, and construct multiple contrastive relationships according to their semantic associations to provide richer self-supervised signals. Extensive experiments and parametric analyzes on eight graph classification real-world datasets well demonstrate the effectiveness of the proposed method.

Graph Neural Networks (GNNs), the powerful graph representation technique based on deep learning, have attracted great research interest in recent years. Although many GNNs have achieved the state-of-the-art accuracy on a set of standard benchmark datasets, they are still limited to traditional semi-supervised framework and lack of sufficient supervision information, especially for the large amount of unlabeled data. To overcome this issue, we propose a novel self-consistent graph neural networks (SCGNN) framework to enrich the supervision information from two aspects: the self-consistency of unlabeled data and the label information of labeled data. First, in order to extract the self-supervision information from the numerous unlabeled nodes, we perform graph data augmentation and leverage a self-consistent constraint to maximize the mutual information of the unlabeled nodes across different augmented graph views. The self-consistency can sufficiently utilize the intrinsic structural attributes of the graph to extract the self-supervision information from unlabeled data and improve the subsequent classification result. Second, to further extract supervision information from scarce labeled nodes, we introduce a fusion mechanism to obtain comprehensive node embeddings by fusing node representations of two positive graph views, and optimize the classification loss over labeled nodes to maximize the utilization of label information. We conduct comprehensive empirical studies on six public benchmark datasets in node classification task. In terms of accuracy, SCGNN improves by an average of 2.08% over the best baseline, and specifically by 5.8% on the Disease dataset.

With rapid worldwide spread of Coronavirus Disease 2019 (COVID-19), jointly identifying severe COVID-19 cases from mild ones and predicting the conversion time (from mild to severe) is essential to optimize the workflow and reduce the clinician's workload. In this study, we propose a novel framework for COVID-19 diagnosis, termed as Structural Attention Graph Neural Network (SAGNN), which can combine the multi-source information including features extracted from chest CT, latent lung structural distribution, and non-imaging patient information to conduct diagnosis of COVID-19 severity and predict the conversion time from mild to severe. Specifically, we first construct a graph to incorporate structural information of the lung and adopt graph attention network to iteratively update representations of lung segments. To distinguish different infection degrees of left and right lungs, we further introduce a structural attention mechanism. Finally, we introduce demographic information and develop a multi-task learning framework to jointly perform both tasks of classification and regression. Experiments are conducted on a real dataset with 1687 chest CT scans, which includes 1328 mild cases and 359 severe cases. Experimental results show that our method achieves the best classification (e.g., 86.86% in terms of Area Under Curve) and regression (e.g., 0.58 in terms of Correlation Coefficient) performance, compared with other comparison methods.

We address the problem of disentangled representation learning with independent latent factors in graph convolutional networks (GCNs). The current methods usually learn node representation by describing its neighborhood as a perceptual whole in a holistic manner while ignoring the entanglement of the latent factors. However, a real-world graph is formed by the complex interaction of many latent factors (e.g., the same hobby, education or work in social network). While little effort has been made toward exploring the disentangled representation in GCNs. In this paper, we propose a novel Independence Promoted Graph Disentangled Networks (IPGDN) to learn disentangled node representation while enhancing the independence among node representations. In particular, we firstly present disentangled representation learning by neighborhood routing mechanism, and then employ the Hilbert-Schmidt Independence Criterion (HSIC) to enforce independence between the latent representations, which is effectively integrated into a graph convolutional framework as a regularizer at the output layer. Experimental studies on real-world graphs validate our model and demonstrate that our algorithms outperform the state-of-the-arts by a wide margin in different network applications, including semi-supervised graph classification, graph clustering and graph visualization.

This paper studies the influence of signal detection schemes on the energy efficiency (EE) of uplink multiple-input-multiple-output (MIMO) systems with low-resolution analog-to-digital converters (ADCs). Assuming equal transmission rates for all users, we derive the optimal power allocation and their analytical approximations for zero-forcing (ZF) and ZF successive interference cancellation (ZF-SIC) receivers. Both the cases with perfect channel state information (CSI) and with imperfect CSI are considered. The EE with different receivers is compared. The results indicate that for uplink massive MIMO systems with low-resolution ADCs, the radio-frequency circuit power consumption can be significant because a large number of antennas are required to compensate for the loss due to quantization errors while the number of base station antennas needed with the ZF-SIC receiver is significantly smaller than that with the ZF receiver. Meanwhile, the increase of power consumption of signal processing with ZF-SIC can be moderate, due to the fact that the receiver weights are reused in a coherent block and the dimensionality is reduced. Consequently, the ZF-SIC receiver is able to improve the overall EE for massive MIMO systems with practical ADCs. We also conduct an approximation analysis for a multi-cell scenario with the pilot contamination and inter-cell-interference considered.

Measuring the intrinsic deformability of arbitrary small‐scale subdivision of a shape is an interesting meanwhile valuable research topic. Such measurement can be directly utilised as a reliable criteria to partition shape into small components and then assist in shape modelling and description. Compared to global modelling, through constructing subdivision‐based complex shape description, the accuracy and flexibility of shape representation can be significantly improved. In this study, the authors propose a line segment advection (LSA)‐based vertex‐level three‐dimensional shape deformability measuring method. It can highlight the deformability characteristics of each shape part in any scale and size. The measurement is realised mainly based on the advection of line segments connecting neighbouring shape mesh vertices. For 3D shapes, since the line segment of triangular mesh facet directly reflects the minimal neighbourhood relationships and mesh microstructure, its advection can capture the finest details of shape deformability. Then, after transferring that information into neighbouring vertices, a vertex‐level shape deformability measurement can be acquired. Besides, to demonstrate the value of the proposed measuring method to shape partitioning and piecewise shape modelling, a straightforward shape partitioning method is introduced as well. Extensive experiments on three publicly available databases are conducted to verify the effectiveness of proposed methods.

Interesting target detection algorithm in complex natural backgrounds images is studied in this paper. Firstly, logGabor filter bank is analyzed, which is consistent with human visual system characteristics. Several kinds of local features from the filter bank can form the integrated feature. Integrated feature congruency (IFC) model is established. And upon compensating noise for IFC, an improved integrated feature congruency (IIFC) model is obtained, in which, target detecting is translated to find the interest points that are significant across scales and orientations. This model is applied to complex natural backgrounds images for target detection. Experimental results show that this method can detect interesting targets effectively from complex natural backgrounds scenes.