Search

In the development of image forensics, detection of Copy-Move Forgery (CMF) has become a major challenge due to the proliferation of image forgery techniques. The CMF is widely utilized to alter the content of the original image to spread false information or to use such forged digital images for illegal purposes e.g. false evidence in the court of law, or to blackmailing any individual. This paper presents a new method for CMF Detection (CMFD) that uses the power of Graph Convolution Networks (GCNs) and its multiple layers with ReLU activation, for CMFD and analysis. The aim to use GCN is due to its ability to improve the feature extraction process by utilizing the spatial and structural affiliation between elements in the digital images. Also, the GCN aims to store information about images and use it to graphically describe images with pixels or image areas as features, spatial and correlation relationships as edges. By pulling data from this image, GCN is able to obtain content rich features that are very powerful at detecting CMF regions. In proposed methodology, we utilized Support Vector machine (SVM) for classification and the binary cross-entropy loss, and the Adam optimizer for improving accuracy. Our scheme successfully achieves high accuracy and is effective in CMFD. We use the MICC F220, and CoMoFoD datasets to test the GCN in our proposed CMFD method. Through much testing and evaluation, we have found that GCN has the tremendous ability for CMFD in digital images in term of accuracy.

Graph Neural Networks (GNNs) offer a promising direction for medical image analysis, particularly due to their ability to capture complex relationships and handle non-Euclidean data structures often encountered in this domain. This study aims to demonstrate the potential of GNNs in medical image classification, showcasing their comparable performance to leading CNNs while using significantly fewer parameters. We introduce a novel GNN model, Graph Convolution Neural Network Enhanced Connectivity (GCNN-EC), incorporating an enhanced connectivity technique that leverages interrelationships between RGB channel features. Evaluated on the MedMNIST dataset, a standardized benchmark for medical image analysis, GCNN-EC performs competitively against pre-trained DNNs, highlighting its potential as an efficient and effective approach for medical image analysis. Importantly, we do not claim that GCNN-EC outperforms all CNNs in every scenario, but rather aim to showcase the potential of GNNs as a valuable tool in the medical imaging toolkit that can be used on its own or in conjuction with CNNs to further advance the state of the art. Additionally, our model outperforms traditional GNNs on standard image datasets like MNIST and CIFAR-10, mitigating the common issue of over-smoothing. These findings underscore the effectiveness of GCNs with dynamic filtering and encourage further research into GNN-based approaches for medical imaging and broader computer vision tasks.

The accuracy of urban traffic flow prediction is influenced by nearby regional road network, historical traffic flow and seasonal climate, which has complex spatial and temporal dependence. In view of the above factors, we proposed a multi-modal traffic flow prediction model fusing road network, historical traffic flow and weather data. Firstly, the weighted spatio-temporal graph was constructed based on the traffic flow time series data, and the weighted STSGCN model was used to extract the spatio-temporal graph features. Secondly, the image sequence was constructed by road network, vehicle track and sensor position data, and the visual features were extracted by ResNet. Finally, based on the MCB and Attention two-channel multi-modal fusion model, the spatio-temporal graph features and the visual features of the image sequence were fused to obtain the aligned fusion vector. Finally, the aligned fusion vector was combined with the weather feature vector to complete the traffic flow prediction. The experimental results showed that the prediction results of our proposed model were better than those of other baseline models. At the same time, the ablation results also proved the effectiveness of each module in our proposed prediction model.

Children with autism spectrum disorder (ASD) often require long-term and high-quality intervention. An important factor affecting the quality of the intervention is therapists’ observation skill, by which therapists can opportunely adjust their intervention strategies based on children’s states. However, there is a shortage of experienced therapists and observational skill development is time-consuming for junior therapists to acquire. This motivates us to use data-driven deep learning method to build an OBservational Therapy-AssIstance Neural network (OBTAIN), which is a weakly-supervised learning framework for the therapy training states recognition. OBTAIN first represents children’s skeleton-sequence data as a large graph. Then a graph representation learning module is used to extract training state features. To learn spatial-temporal behavior features more effectively and efficiently, a novel structure-aware GCN block is designed in OBTAIN’s learning module. Finally, a MILnet and corresponding joint MIL loss are used to learn state score prediction from extracted features. Experimental results (0.824 AUC score) on DREAM dataset demonstrate our OBTAIN can effectively recognize the training states in autism intervention.

Graph convolutional neural networks (GCNs) are emerging neural networks for graph structures that include large features associated with each vertex. The operations of GCN can be divided into two phases - aggregation and combination. While the combination just performs matrix multiplications using trained weights and aggregated features, the aggregation phase requires graph traversal to collect features from adjacent vertices. Even though neural network applications rely on GPU’s massively parallel processing, GCN aggregation kernels exhibit rather low performance since graph processing using compressed graph structures provokes frequent irregular accesses in GPUs. In order to investigate the performance hurdles of GCN aggregation on GPU, we perform an in-depth analysis of the aggregation kernels using real GPU hardware and a cycle-accurate GPU simulator. We first analyze the characteristics of the popular graph datasets used for GCN studies. We reveal the fractions of non-zero elements in feature vectors are diverse among datasets. Based on the observation, we build two types of aggregation kernels that handle uncompressed and compressed feature vectors. Our evaluation exhibits the performance of aggregation can be significantly influenced by kernel design approaches and feature density. We also analyze the individual loads that access the data arrays of the aggregation kernels to specify critical loads. Our analysis reveals the performance of GPU memory hierarchy is influenced by access patterns and feature size of graph datasets. Based on our observations we discuss possible kernel design approaches and architectural ideas that can improve the performance of GCN aggregation.

In order to automatically generate multiple labels for the event text of the Chinese government hotline, this paper propose a multi-label classification framework based on graph convolutional network (GCN), BERT, and memory network. The framework consists of three modules: label count prediction module, label semantic insert module, and label selection module. In the label count prediction module, this paper constructs the event graph with the abstract meaning representation (AMR) and extract the event topic information vector with GCN. To predict the label count, this paper first use BERT to extract the event semantic information vector and then fuse it with the event topic information vector (GCN-BERT fusion vector) with a dynamic fusion gate. In the label semantic insert module, to obtain the event label candidate set, this paper uses a multi-hop memory network to store the event label semantic information, and then use the answer selection framework, which matches the GCN-BERT fusion vector with the event label semantic memory vector. In label selection module, this paper uses the label count based multi-label selection to sort the event label candidate set and guide to output the optimal multi-label set of the event. Comparison experimental results show that the proposed framework outperforms all baselines and ablation studies demonstrate the effectiveness of each module.

Context-based GCNs have achieved relatively good effectiveness in the sentiment analysis task, especially aspect-level sentiment analysis (ALSA). However, the previous context-based GCNs for ALSA often used GCNs with the following limitations: (i) Using GCNs limited to a few layers (two or three) due to the vanishing gradient, limiting their performance. (ii) Not considering helpful information about the hidden context between the words. To solve these limitations, this paper proposes a novel CNN over the BERT-GCN model for ALSA. The contributions of the proposed method are summarized as follows: (i) Handling the disadvantage of limiting the GCN to a few layers by adding convolutional layers of the convolutional neural network (CNN) model after GCN layers. (ii) Considering further helpful information about the hidden context between the words by combining the Bidirectional Encoder Representations from Transformers (BERT) and Bidirectional Long Short Term Memory (BiLSTM) models. The proposed model includes the following steps: First, words in sentences are converted vectors using BERT. Second, the contextualized word representations are created based on BiLSTM over word vectors. Third, significant features are extracted and represented using the GCN model with multiple convolutional layers over the contextualized word representations. Finally, the aspect-level sentiments are classified using the CNN model over the feature vectors. Experiments on three benchmark datasets illustrate that our proposed model has improved the performance of the previous context-based GCN methods for ALSA.

The problem of image segmentation is one of the most significant ones in computer vision. Recently, deep-learning methods have dominated state-of-the-art solutions that automatically or interactively divide an image into subregions. However, the limitation of deep-learning approaches is that they require a substantial amount of training data, which is costly to prepare. An alternative solution is semi-supervised image segmentation. It requires rough denotations to define constraints that are next generalized to precisely delimit relevant image regions without using train examples. Among semi-supervised strategies for image segmentation, the leading are graph-based techniques that define image segmentation as a result of pixel or region affinity graph partitioning. This paper revisits the problem of graph-based image segmentation. It approaches the problem as semi-supervised node classification in the SLIC superpixels region adjacency graph using a graph convolutional network (GCN). The performance of both spectral and spatial graph convolution operators is considered, represented by Chebyshev convolution operator and GraphSAGE respectively. The results of the proposed method applied to binary and multi-label segmentation are presented, numerically assessed, and analyzed. In its best variant, the proposed method scored the average DICE of 0.86 in the binary segmentation task and 0.79 in the multi-label segmentation task. Comparison with state-of-the-art graph-based methods, including Random Walker and GrabCut, shows that graph convolutional networks can represent an attractive alternative to the existing solutions to graph-based semi-supervised image segmentation.

With the development of big data, large-scale traffic flow forecasting which is a part of smart transportation has become an increasingly important research direction. Accurate and real-time traffic flow prediction is the key and difficult part of the traffic. The complex spatial topological structure and dynamic traffic flow information in urban roads constitute a changeable spatial correlation, and the daily traffic flow cycle and weekly traffic flow cycle constitute a complex time correlation. For the current mainstream model, there are two main limitations: 1. Most of the existing models only focus on time correlation and ignore spatial correlation. 2. Even if the spatial correlation is concerned, the topological relationship between spaces is not fully considered. This paper proposes a new traffic-flow prediction model, which named Principal Spatio-Temporal Graph Convolution Network (PST-GCN) model, which uses a combination of Principal Component Analysis (PCA), Graph Convolution Network (GCN), and Long Short-Term Memory model (LSTM). Specifically, PCA is used to reduce the dimension of data, GCN is used to learn the network topology of urban roads, LSTM is used to capture the time correlation of traffic flow. By comparing the results of different models, the proposed model is better than the current mainstream models.

We propose SiCaGCN, a learning system to predict the similarity of a given software code to a set of codes that are permitted to run on a computational resource, such as a supercomputer or a cloud server. This code characterization allows us to detect abusive codes. Our system relies on a structural analysis of the control-flow graph of the software codes and two different graph similarity measures: Graph Edit Distance (GED) and a singular values based metric. SiCaGCN combines elements of Graph Convolutional Neural Networks (GCN), Capsule networks, attention mechanism, and neural tensor networks. Our experimental results include a study of the trade-offs between the two similarity metrics and two variations of our learning networks, with and without the use of capsules. Our main findings are that the use of capsules reduces mean square error significantly for both similarity metrics. Use of capsules reduces the runtime to calculate the GED while increases the runtime of singular values calculation.

Human video prediction is still a challenging problem due to the uncertainty of future actions and complexity of frame details. Recent methods tackle this problem in two steps: firstly to forecast future human poses from the initial ones, and then to generate realistic frames conditioned on predicted poses. Following this framework, we propose a novel Graph Convolutional Network (GCN) based pose predictor to comprehensively model human body joints and forcast their positions holistically, and also a stacked generative model with a temporal discriminator to iteratively refine the quality of the generated videos. The GCN based pose predictor fully considers the relationships among body joints and produces more plausible pose predictions. With the guidance of predicted poses, a temporal discriminator encodes temporal information into future frame generation to achieve high-quality results. Furthermore, stacked residual refinement generators make the results more realistic. Extensive experiments on benchmark datasets demonstrate that the proposed method produces better predictions than state-of-the-arts and achieves up to 15% improvement in PSNR.