Learning Convolutional Network Research Articles

Beyond pixel: Superpixel-based MRI segmentation through traditional machine learning and graph convolutional network

Background and Objective:Tendon segmentation is crucial for studying tendon-related pathologies like tendinopathy, tendinosis, etc. This step further enables detailed analysis of specific tendon regions using automated or semi-automated methods. This study specifically aims at the segmentation of Achilles tendon, the largest tendon in the human body. Methods:This study proposes a comprehensive end-to-end tendon segmentation module composed of a preliminary superpixel-based coarse segmentation preceding the final segmentation task. The final segmentation results are obtained through two distinct approaches. In the first approach, the coarsely generated superpixels are subjected to classification using Random Forest (RF) and Support Vector Machine (SVM) classifiers to classify whether each superpixel belongs to a tendon class or not (resulting in tendon segmentation). In the second approach, the arrangements of superpixels are converted to graphs instead of being treated as conventional image grids. This classification process uses a graph-based convolutional network (GCN) to determine whether each superpixel corresponds to a tendon class or not. Results:All experiments are conducted on a custom-made ankle MRI dataset. The dataset comprises 76 subjects and is divided into two sets: one for training (Dataset 1, trained and evaluated using leave-one-group-out cross-validation) and the other as unseen test data (Dataset 2). Using our first approach, the final test AUC (Area Under the ROC Curve) scores using RF and SVM classifiers on the test data (Dataset 2) are 0.992 and 0.987, respectively, with sensitivities of 0.904 and 0.966. On the other hand, using our second approach (GCN-based node classification), the AUC score for the test set is 0.933 with a sensitivity of 0.899. Conclusions:Our proposed pipeline demonstrates the efficacy of employing superpixel generation as a coarse segmentation technique for the final tendon segmentation. Whether utilizing RF, SVM-based superpixel classification, or GCN-based classification for tendon segmentation, our system consistently achieves commendable AUC scores, especially the non-graph-based approach. Given the limited dataset, our graph-based method didn’t perform as well as non-graph-based superpixel classifications; however, the results obtained provide valuable insights into how well the models can distinguish between tendons and non-tendons. This opens up opportunities for further exploration and improvement.

Vision transformer and CNN-based skin lesion analysis: classification of monkeypox

AbstractMonkeypox is an important health problem. Rapid diagnosis of monkeypox skin lesions and emergency isolation when necessary is essential. Also, some skin lesions, such as melanoma, can be fatal and must be rapidly distinguished. However, in some cases, it is difficult to distinguish the lesions visually. Methods such as dermoscopy, high-resolution ultrasound imaging, etc. can be used for better observation. But these methods are often based on qualitative analysis, subjective and time-consuming. Therefore, in this study, a quantitative and objective classification tool has been developed to assist dermatologists and scientists. The proposed system classifies seven skin lesions, including monkeypox. A popular approach Vision Transformer and some popular deep learning convolutional networks have been trained with the transfer learning approach and all results have been compared. Then, the models that show the best accuracy score have been combined to make the final prediction using bagging-ensemble learning. The proposed ensemble-based system produced 81.91% Accuracy, 65.94% Jaccard, 87.16% Precision, 74.12% Recall, and 78.16% Fscore values. In terms of different criteria metrics, the system produced competitive or even better results than the literature.

Data-Driven Prediction Model for Analysis of Sensor Data

In view of Industry 4.0, data generation and analysis are challenges. For example, machine health monitoring and remaining useful life prediction use sensor signals, which are difficult to analyze using traditional methods and mathematical techniques. Machine and deep learning algorithms have been used extensively in Industry 4.0 to process sensor signals and improve the accuracy of predictions. Therefore, this paper proposes and validates the data-driven prediction model to analyze sensor data, including in the data transformation phase Principal Component Analysis tested by Fourier Transformation and Wavelet Transformation, and the modeling phase based on machine and deep learning algorithms. The machine learning algorithms used for tests in this research are Random Forest Regression (RFR), Multiple Linear Regression (MLR), and Decision Tree Regression (DTR). For the deep learning comparison, the algorithms are Deep Learning Regression and Convolutional network with LeNet-5 Architecture. The experimental results indicate that the models show promising results in predicting wear values and open the problem to further research, reaching peak values of 92.3% accuracy for the first dataset and 62.4% accuracy for the second dataset.

Detecting health misinformation: A comparative analysis of machine learning and graph convolutional networks in classification tasks

In the digital age, the proliferation of health-related information online has heightened the risk of misinformation, posing substantial threats to public well-being. This research conducts a meticulous comparative analysis of classification models, focusing on detecting health misinformation. The study evaluates the performance of traditional machine learning models and advanced graph convolutional networks (GCN) across critical algorithmic metrics. The results comprehensively understand each algorithm's effectiveness in identifying health misinformation and provide valuable insights for combating the pervasive spread of false health information in the digital landscape. GCN with TF-IDF gives the best result, as shown in the result section.•The research method involves a comparative analysis of classification algorithms to detect health misinformation, exploring traditional machine learning models and graph convolutional networks.•This research used algorithms such as Passive Aggressive Classifier, Random Forest, Decision Tree, Logistic Regression, Light GBM, GCN, GCN with BERT, GCN with TF-IDF, and GCN with Word2Vec were employed. Performance Metrics: Accuracy: for Passive Aggressive Classifier: 85.75 %, Random Forest: 86 %, Decision Tree: 81.30 %, Light BGM: 83.29 %, normal GCN: 84.53 %, GCN with BERT: 85.00 %, GCN with TR-IDF: 93.86 % and GCN with word2Vec: 81.00 %•Algorithmic performance metrics, including accuracy, precision, recall, and F1-score, were systematically evaluated to assess the efficacy of each model in detecting health misinformation, focusing on understanding the strengths and limitations of different approaches. The superior performance of Graph Convolutional Networks (GCNs) with TF-IDF embedding, achieving an accuracy of 93.86 %

Formation energy prediction of crystalline compounds using deep convolutional network learning on voxel image representation

Emerging machine-learned models have enabled efficient and accurate prediction of compound formation energy, with the most prevalent models relying on graph structures for representing crystalline materials. Here, we introduce an alternative approach based on sparse voxel images of crystals. By developing a sophisticated network architecture, we showcase the ability to learn the underlying features of structural and chemical arrangements in inorganic compounds from visual image representations, subsequently correlating these features with the compounds’ formation energy. Our model achieves accurate formation energy prediction by utilizing skip connections in a deep convolutional network and incorporating augmentation of rotated crystal samples during training, performing on par with state-of-the-art methods. By adopting visual images as an alternative representation for crystal compounds and harnessing the capabilities of deep convolutional networks, this study extends the frontier of machine learning for accelerated materials discovery and optimization. In a comprehensive evaluation, we analyse the predicted convex hulls for 3115 binary systems and introduce error metrics beyond formation energy error. This evaluation offers valuable insights into the impact of formation energy error on the performance of the predicted convex hulls.

Identification of epileptic networks with graph convolutional network incorporating oscillatory activities and evoked synaptic responses

Stereoelectroencephalography (SEEG) offers unique neural data from in-depth brain structures with fine temporal resolutions to better investigate the origin of epileptic brain activities. Although oscillatory patterns from different frequency bands and functional connectivity computed from the SEEG datasets are employed to study the epileptic zones, direct electrical stimulation-evoked electrophysiological recordings of synaptic responses, namely cortical-cortical evoked potentials (CCEPs), from the same SEEG electrodes are not explored for the localization of epileptic zones. Here we proposed a two-stream model with unsupervised learning and graph convolutional network tailored to the SEEG and CCEP datasets in individual patients to perform localization of epileptic zones. We compared our localization results with the clinically marked electrode sites determined for surgical resections. Our model had good classification capability when compared to other state-of-the-art methods. Furthermore, based on our prediction results we performed group-level brain-area mapping analysis for temporal, frontal and parietal epilepsy patients and found that epileptic and non-epileptic brain networks were distinct in patients with different types of focal epilepsy. Our unsupervised data-driven model provides personalized localization analysis for the epileptic zones. The epileptic and non-epileptic brain areas disclosed by the prediction model provide novel insights into the network-level pathological characteristics of epilepsy.

Short‐term photovoltaic power forecasting using hybrid contrastive learning and temporal convolutional network under future meteorological information absence

AbstractPhotovoltaic (PV) power generation is widely utilized to satisfy the increasing energy demand due to its cleanness and inexhaustibility. Accurate PV power forecasting can improve the penetration of PV power in the grid. However, it is pretty challenging to predict PV power in short‐term under precious future meteorological information absence conditions. To address this problem, this study proposes the hybrid Contrastive Learning and Temporal Convolutional Network (CL‐TCN), and this forecasting approach consists of two parts, including model training and adaptive processes of forecasting models. In the model training stage, this forecasting method firstly trains 18 TCN models for 18 time points from 9:00 a.m. to 17:30 p.m. These TCN models are trained by only using historical PV power data samples, and each model is used to predict the next half‐hour power output. The adaptive process of models means that, in a practical forecasting stage, PV power samples from historical data are firstly evaluated and scored by a CL based data scoring mechanism to search for the most similar data samples to current measured samples. Then these similar samples are further applied to training a single above‐mentioned well‐trained TCN model to improve its performance in forecasting the next half‐hour PV power. The experimental results tested at the time resolution of 30 min demonstrate that the proposed approach has superior performance in forecasting accuracy not only in smooth PV power samples but also in fluctuating PV power samples. Moreover, the proposed CL based data scoring mechanism can filter useless data samples effectively accelerating the forecasting process.

Deep image clustering with contrastive learning and multi-scale graph convolutional networks

Deep clustering has shown its promising capability in joint representation learning and clustering via deep neural networks. Despite the significant progress, the existing deep clustering works mostly utilize some distribution-based clustering loss, lacking the ability to unify representation learning and multi-scale structure learning. To address this, this paper presents a new deep clustering approach termed Image clustering with contrastive learning and multi-scale Graph Convolutional Networks (IcicleGCN), which bridges the gap between convolutional neural network (CNN) and graph convolutional network (GCN) as well as the gap between contrastive learning and multi-scale structure learning for the deep clustering task. Our framework consists of four main modules, namely, the CNN-based backbone, the Instance Similarity Module (ISM), the Joint Cluster Structure Learning and Instance reconstruction Module (JC-SLIM), and the Multi-scale GCN module (M-GCN). Specifically, the backbone network with two weight-sharing views is utilized to learn the representations for the two augmented samples (from each image). The learned representations are then fed to ISM and JC-SLIM for joint instance-level and cluster-level contrastive learning, respectively, during which an auto-encoder in JC-SLIM is also pretrained to serve as a bridge to the M-GCN module. Further, to enforce multi-scale neighborhood structure learning, two streams of GCNs and the auto-encoder are simultaneously trained via (i) the layer-wise interaction with representation fusion and (ii) the joint self-adaptive learning. Experiments on multiple image datasets demonstrate the superior clustering performance of IcicleGCN over the state-of-the-art. The code is available at https://github.com/xuyuankun631/IcicleGCN.

Face and number plate recognition for car anti-theft

Smart parking systems along with that is continuous development of new technologies, are widely applied to improve our lives. It can also add new technologies with advanced functions making it a multi-functional management system. Thanks to the anti-theft technologies that are installed, modern cars are significantly more difficult to steal than they once were. However, electrical systems can still experience issues, though and malfunction at some point. This paper suggests using video image recognition technology, at car parks and parking lots as an anti-theft solution, alerting the presence of the non-owner of vehicle in the driver’s seat. The system automatically predicts whether the driver is valid with the registration number plate. The image of the car is captured by the camera at the entrance gates of the parking lot. The proposed algorithm includes face recognition in images, building a deep learning convolutional network that classifies faces (subscriber’s images); using Cascade trainer to train number plate object recognition, vehicle number recognition through character recognition technique. The system can recognize reality through a personal computer connected to the camera at the scene or through photos and video files. Result, the model can face recognition and match to license plate in at a moment.

Class-aware progressive self-training for learning convolutional networks on graphs

Learning convolutional networks on graphs have been a popular topic for machine learning on graph-structured data and achieved state-of-the-art results on various practical tasks. However, most existing works ignore the impact of per-class distribution, therefore their performance may be limited due to the diversity of various categories. In this paper, we propose a novel class-aware progressive self-training (CPS) algorithm for training graph convolutional networks (GCNs). Compared to other self-training algorithms for GCNs’ learning, the proposed CPS algorithm leverages the class distribution to update the original graph structure in each self-training loop, including: (a) find these high-confident unlabeled nodes in the graph for each category to add pseudo labels, in order to enlarge the current set of labeled nodes; (b) delete these noisy edges between different classes for graph sparsification. Then, the optimized graph is used for next self-training loops in hopes of enhancing the classification performance. We evaluate the proposed CPS on several datasets commonly used for GCNs’ learning, and the experimental results show that the proposed CPS algorithm outperforms other baselines.

Learning Large Neighborhood Search for Vehicle Routing in Airport Ground Handling

Dispatching vehicle fleets to serve flights is a key task in airport ground handling (AGH). Due to the notable growth of flights, it is challenging to simultaneously schedule multiple types of operations (services) for a large number of flights, where each type of operation is performed by one specific vehicle fleet. To tackle this issue, we first represent the operation scheduling as a complex vehicle routing problem and formulate it as a mixed integer linear programming (MILP) model. Then given the graph representation of the MILP model, we propose a learning assisted large neighborhood search (LNS) method using data generated based on real scenarios, where we integrate imitation learning and graph convolutional network (GCN) to learn a destroy operator to automatically select variables, and employ an off-the-shelf solver as the repair operator to reoptimize the selected variables. Experimental results based on a real airport show that the proposed method allows for handling up to 200 flights with 10 types of operations simultaneously, and outperforms state-of-the-art methods. Moreover, the learned method performs consistently accompanying different solvers, and generalizes well on larger instances, verifying the versatility and scalability of our method.

Physical-Layer Secret Key Generation Based on Bidirectional Convergence Feature Learning Convolutional Network

Physical layer secret key generation (PLKG) is a new research area that has emerged in recent years. It is aimed at scenarios where legitimate IoT devices communicate directly, interacting with confidential information for lower overhead and higher security by using wireless channel. When applying it to wireless feature extraction, noise removal is not taken into account in current deep learning networks. To address these problems, the PLKG scheme based on bidirectional convergence feature learning convolutional network (BCFL-based scheme) is proposed, which consists of neural network called BCFL and a new quantization method to achieve better secret key generation. Unlike existing PLKG schemes that enabling both parties to communicate for obtaining higher channel feature similarities, when training it, channel state information (CSI) obtained by channel estimation for two legitimate devices during coherent time is used as inputs; and mean square error (MSE) between two outputs is used as result of loss function for iterative training. Thus, it can obtain better denoising ability with guaranteed low computational resource consumption, and two legitimate devices can obtain highly correlated channel features. Multiple quantization method is also proposed to address low secret key generation rate (KGR) and low secret key randomness (KR). The results show that the proposed BCFL-based scheme has a lower MSE than other schemes in different scenarios, indicating that it has better capability to learn channel reciprocity; and secret key error rate (KER) and time consumption are only about 50% of other schemes, which is a significant performance improvement.

DL-PCN: Differential learning and parallel convolutional network for action recognition

Graph Convolution Network (GCN) algorithms have greatly improved the accuracy of skeleton-based human action recognition. GCN can utilize the spatial information between skeletal joints in subsequent frames better than other deep learning algorithms, which is beneficial for achieving high accuracy. However, the traditional GCN algorithms consume lots of computation for the stack of multiple primary GCN layers. Aiming at solving the problem, we introduce a lightweight network, a Differential Learning and Parallel Convolutional Networks (DL-PCN), whose key modules are Differential Learning (DLM) and the Parallel Convolutional Network (PCN). DLM features a feedforward connection, which carries the error information of GCN modules with the same structure, where GCN and CNN modules directly extract the original information from the input data, making the spatiotemporal information extracted by these modules more complete than that of GCN and CNN tandem structure. PCN comprises GCN and Convolution Neural Network (CNN) in parallel. Our network achieves comparable performance on the NTU RGB+D 60 dataset, the NTU RGB+D 120 dataset and the Northwestern-UCLA dataset while considering both accuracy and calculation parameters.

FedAGCN: A traffic flow prediction framework based on federated learning and Asynchronous Graph Convolutional Network

FedAGCN: A traffic flow prediction framework based on federated learning and Asynchronous Graph Convolutional Network

Deep Learning Convolutional Network for Bimodal Biometric Recognition with Information Fusion at Feature Level

Biometric recognition has been an extensively researched field in recent years due to the growth of its applications in daily activities. State of the art work in biometrics proposes the implementation of multimodal systems that employ one or more traits to increase the security of the system since it is more difficult for an impostor to acquire, falsify or forge multiple samples of different traits from an enrolled user. In this paper, we propose the implementation of a Deep Learning bimodal network that combines voice and face modalities. Voice feature extraction was done with a SincNet arquitecture and face image features were extracted with a set of convolutional layers. The feature vectors of both modalities are combined within the network with two methods: averaging or concatenation. The averaged/concatenated vector is further processed with a fully connected layer to output a bimodal vector that contains discriminatory information of an individual. The bimodal vector is used with a fully connected layer with the softmax function to perform the identification task. The verification task is performed by matching the bimodal vector with a template to obtain a score that must be used to either accept or reject an users identity. We compared the results yielded by both fusion methods implemented in our proposed network for both recognition tasks. Both methods achieved an accuracy as high as 99 % in the identification task and an Equal Error Rate (EER) as low as 0.14 % for verification. These results were obtained by combining BIOMEX-DB and VidTimit databases.

Extracting topological features to identify at-risk students using machine learning and graph convolutional network models

Technological advances have significantly affected education, leading to the creation of online learning platforms such as virtual learning environments and massive open online courses. While these platforms offer a variety of features, none of them incorporates a module that accurately predicts students’ academic performance and commitment. Consequently, it is crucial to design machine learning (ML) methods that predict student performance and identify at-risk students as early as possible. Graph representations of student data provide new insights into this area. This paper describes a simple but highly accurate technique for converting tabulated data into graphs. We employ distance measures (Euclidean and cosine) to calculate the similarities between students’ data and construct a graph. We extract graph topological features (GF) to enhance our data. This allows us to capture structural correlations among the data and gain deeper insights than isolated data analysis. The initial dataset (DS) and GF can be used alone or jointly to improve the predictive power of the ML method. The proposed method is tested on an educational dataset and returns superior results. The use of DS alone is compared with the use of DS + GF in the classification of students into three classes: “failed”,“at risk”, and “good”. The area under the receiver operating characteristic curve (AUC) reaches 0.948 using DS, compared with 0.964 for DS + GF. The accuracy in the case of DS + GF varies from 84.5 to 87.3%. Adding GF improves the performance by 2.019% in terms of AUC and 3.261% in terms of accuracy. Moreover, by incorporating graph topological features through a graph convolutional network (GCN), the prediction performance can be enhanced by 0.5% in terms of accuracy and 0.9% in terms of AUC under the cosine distance matrix. With the Euclidean distance matrix, adding the GCN improves the prediction accuracy by 3.7% and the AUC by 2.4%. By adding graph embedding features to ML models, at-risk students can be identified with 87.4% accuracy and 0.97 AUC. The proposed solution provides a tool for the early detection of at-risk students. This will benefit universities and enhance their prediction performance, improving both effectiveness and reputation.

ResFCNET: A Skin Lesion Segmentation Method Based on a Deep Residual Fully Convolutional Neural Network

Melanoma, a high-level variant of skin cancer is very difficult to distinguish from other skin cancer types in patients. The presence of large variety of sizes of lesions, fuzzy boundaries and irregular shaped nature, with low contrast between skin lesions and surrounding fresh areas makes it clinically difficult to detect and treat melanoma. In this paper, we propose Residual Full Convolutional Network (ResFCNET) skin lesion recognition model that combines residual learning and full convolutional network to perform semantic segmentation of skin lesion. Based on secondary feature extraction and classification, experiment was done to verify the effectiveness of our model using ISBI 2016 and ISBI 2017 dataset. Results showed that residual convolution neural network obtain high precision classification. This technique is novel and provides a compelling insight for medical image segmentation.

Siamese labels auxiliary learning

In deep learning, auxiliary modules for model training have become increasingly popular, such as Deep Mutual Learning (DML) and Multi-Scale Dense Convolutional Networks (MSDNet), which can maximize the performance of the model without increasing the amount of test computation. Nevertheless, current research does not fully exploit the knowledge between different auxiliary modules. This paper proposes a new model training technique–Siamese Labels Auxiliary (SiLA) Learning, in which the SiLA module is designed to concatenate the outputs of the modules to get auxiliary information. In addition, the knowledge learned from different modules can be fully exploited by auxiliary information. Our experiments show that SiLA Learning effectively improves the performance of standard models and achieves convincing experimental results on image classification tasks without increasing the amount of test computation. Moreover, the experiments show that SiLA Learning can be easily combined with methods such as DML and MSDNet to exploit the knowledge fully and perform best.

Joint Threshold Learning Convolutional Networks for Intelligent Fault Diagnosis Under Nonstationary Conditions

Joint Threshold Learning Convolutional Networks for Intelligent Fault Diagnosis Under Nonstationary Conditions

MF-GCN: Motion Flow-Based Graph Network Learning Dynamics for Aerial IR Target Recognition

The primary technology of the infrared imaging seeker optoelectronic countermeasure system includes the automatic recognition and tracking of air targets. New complex infrared interference factors pose severe challenges to the ability of infrared thermal imaging seekers to identify and lock targets accurately. Despite significant recent advances, machine vision systems still lag substantially behind biological vision systems in terms of performance and robustness. The distinguishing feature of the brain is its ability to detect learned objects under various non-ideal situations. Inspired by the sensitivity of the dorsal visual pathway to object motion information, a graph convolutional network learning model is developed for aerial target recognition, called the adaptive graph reasoning method driven by motion flow model, which employs the predicted object optical flow information to promote the keypoint estimation performance. First, an adaptive graph fusion module is designed to establish an adaptation mechanism to generalize keypoint features from the regular network to the target skeleton graph model. Then, the optical flow estimation features are utilized to understand the graph reasoning of the target keypoint motion. Finally, we propose a aerial target pose estimation framework which effectively integrates the temporal cues across frames by combining the motion features of optical flow estimation, and effectively promotes the estimation accuracy of keypoints. Finally, our extensive experiments in the established aerial target flight infrared dataset show the efficiency of the presented model.