Siamese Convolutional Neural Network Research Articles

Recognition of autism in subcortical brain volumetric images using autoencoding-based region selection method and Siamese Convolutional Neural Network

Recognition of autism in subcortical brain volumetric images using autoencoding-based region selection method and Siamese Convolutional Neural Network

Remote Sensing Building Damage Assessment Based on Machine Learning

Abstract After the occurrence of various types of disasters, including natural disasters and man-made damage, aid workers need accurate and timely data, such as the damage status of buildings, in order to take effective measures for rescue. So as to solve this problem, this paper researches and designs a building damage classification system based on machine learning. The damage assessment system consists of two network models (building extraction network and damage classification network). This article analyzes and designs the structure of each network model, and discusses the principles related to computer vision in machine learning. Buildings in satellite images are segmented through Siamese Convolutional Neural Network, the BottleNeck Module and Feature Pyramid Network are used in the damage classification assessment network to detect damage to buildings in sub-temporal remote sensing images. Subsequently, the model was trained and tested on different disaster events on the xBD dataset. The results show that the building damage detection system based on Siamese-CNN achieves good detection accuracy, and the system has the advantages of simple operation, good timeliness and low resource consumption, and can well meet the needs of disaster assessment.

A Peak-Finding Siamese Convolutional Neural Network (PF-SCNN) for Aero-Engine Hot Jet FT-IR Spectrum Classification

Aiming at solving difficulties related to aero-engine classification and identification, two telemetry Fourier transform infrared spectrometers are utilized to measure the infrared spectra of six types of aero-engine hot jets, and create a spectral data set, which is divided into a training set (80%), a validation set (10%), and a prediction set (10%). A peak-finding Siamese convolutional neural network (PF-SCNN) is used to match and classify the spectral data. During the training stage, the Siamese convolutional neural network (SCNN) is designed to extract spectral features and calculate the distance similarity. In order to improve the efficiency of the SCNN, a peak-finding method is introduced to extract the spectral peaks, which are used to train the model instead of the original spectral data. During the prediction stage, the trained model is used to calculate the similarity between the prediction set and the combined set of the training set and validation set, and the label of the most similar training data in each prediction set is used as the prediction label. The performance measures of the classification results include accuracy, precision, recall, confusion matrix, and F1-score. The experimental results show that the PF-SCNN can achieve a high classification accuracy rate of 99% and can complete the task of classifying the infrared spectra of aero-engine hot jets.

Inter-Turn Short Circuit Diagnosis of Permanent Magnet Synchronous Motor Based on Siamese Convolutional Neural Network Under Small Fault Samples

Inter-Turn Short Circuit Diagnosis of Permanent Magnet Synchronous Motor Based on Siamese Convolutional Neural Network Under Small Fault Samples

Identification of quantum entanglement with Siamese convolutional neural networks and semisupervised learning

Identification of quantum entanglement with Siamese convolutional neural networks and semisupervised learning

SURE: A Visualized Failure Indexing Approach using Program Memory Spectrum

Failure indexing is a longstanding crux in software debugging, the goal of which is to automatically divide failures (e.g., failed test cases) into distinct groups according to the culprit root causes, as such multiple faults residing in a faulty program can be handled independently and simultaneously. The community of failure indexing has long been plagued by two challenges: 1) The effectiveness of division is still far from promising. Specifically, existing failure indexing techniques only employ a limited source of software run-time data, for example, code coverage, to be failure proximity and further divide them, which typically delivers unsatisfactory results. 2) The outcome can be hardly comprehensible. Specifically, a developer who receives the division result is just aware of how all failures are divided, without knowing why they should be divided the way they are. This leads to difficulties for developers to be convinced by the division result, which in turn affects the adoption of the results. To tackle these two problems, in this paper, we propose SURE, a vi SU alized failu R e ind E xing approach using the program memory spectrum. We first collect the run-time memory information (i.e., variables’ names and values, as well as the depth of the stack frame) at several preset breakpoints during the execution of a failed test case, and transform the gathered memory information into a human-friendly image (called program memory spectrum, PMS). Then, any pair of PMS images that serve as proxies for two failures is fed to a trained Siamese convolutional neural network, to predict the likelihood of them being triggered by the same fault. Last, a clustering algorithm is adopted to divide all failures based on the mentioned likelihood. In the experiments, we use 30% of the simulated faults to train the neural network, and use 70% of the simulated faults as well as real-world faults to test. Results demonstrate the effectiveness of SURE: It achieves 101.20% and 41.38% improvements in faults number estimation, as well as 105.20% and 35.53% improvements in clustering, compared with the state-of-the-art technique in this field, in simulated and real-world environments, respectively. Moreover, we carry out a human study to quantitatively evaluate the comprehensibility of PMS, revealing that this novel type of representation can help developers better comprehend failure indexing results.

A Siamese deep learning framework for efficient hardware Trojan detection using power side-channel data

Hardware Trojans (HTs) are hidden threats embedded in the circuitry of integrated circuits (ICs), enabling unauthorized access, data theft, operational disruptions, or even physical harm. Detecting Hardware Trojans (HTD) is paramount for ensuring IC security. This paper introduces a novel Siamese neural network (SNN) framework for non-destructive HTD. The proposed framework can detect HTs by processing power side-channel signals without the need for a golden model of the IC. To obtain the best results, different neural network models such as Convolutional Neural Network (CNN), Gated Recurrent Unit (GRU), and Long Short-Term Memory (LSTM) are integrated individually with SNN. These models are trained on the extracted features from the Trojan Power & EM Side-Channel dataset. The results show that the Siamese LSTM model achieved the highest accuracy of 86.78%, followed by the Siamese GRU model with 83.59% accuracy and the Siamese CNN model with 73.54% accuracy. The comparison shows that of the proposed Siamese LSTM is a promising new approach for HTD and outperform the state-of-the-art methods.

A New Method for Bearing Fault Diagnosis across Machines Based on Envelope Spectrum and Conditional Metric Learning.

In recent years, most research on bearing fault diagnosis has assumed that the source domain and target domain data come from the same machine. The differences in equipment lead to a decrease in diagnostic accuracy. To address this issue, unsupervised domain adaptation techniques have been introduced. However, most cross-device fault diagnosis models overlook the discriminative information under the marginal distribution, which restricts the performance of the models. In this paper, we propose a bearing fault diagnosis method based on envelope spectrum and conditional metric learning. First, envelope spectral analysis is used to extract frequency domain features. Then, to fully utilize the discriminative information from the label distribution, we construct a deep Siamese convolutional neural network based on conditional metric learning to eliminate the data distribution differences and extract common features from the source and target domain data. Finally, dynamic weighting factors are employed to improve the convergence performance of the model and optimize the training process. Experimental analysis is conducted on 12 cross-device tasks and compared with other relevant methods. The results show that the proposed method achieves the best performance on all three evaluation metrics.

Two-stage prediction method for capacity aging trajectories of lithium-ion batteries based on Siamese-convolutional neural network

Accurate battery lifespan prediction can guarantee battery safety in applications. Considering the complex nonlinear nature of battery degradation, it is more challenging to predict battery aging trajectories utilizing less data. A two-stage prediction method for battery capacity aging trajectories is proposed. The first stage is battery lifespan prediction based on the Siamese-Convolutional neural network (Siamese-CNN). Taking the predicted lifespan as its prior information, the second stage is predicting capacity aging trajectories with the Convolutional neural network (CNN). The method only requires partial voltage-current data from 30 cycles, of which the similarity in different batteries is obtained by the Siamese-CNN. The lifespan-unknown batteries in the testing set are fed into the Siamese-CNN to match the lifespan-known training set battery with the highest similarity. Then, CNN is employed to predict the capacity aging trajectories under different profiles. The ratio of each cell's predicted lifespan to the fixed length of the CNN output sequence is regarded as the interval for uniform sampling to keep the series consistent. The prediction accuracy of the method is validated by an open dataset. It is demonstrated that the battery lifespan and capacity aging trajectory prediction achieve a mean absolute percentage error (MAPE) of 2.44% and 1.28%, respectively.

Predicting diabetic retinopathy stage using Siamese Convolutional Neural Network

ABSTRACT Diabetic retinopathy (DR) is considered to be a leading cause of blindness in people aged 16 to 64 years, affecting around 40% of the population diagnosed with diabetes mellitus (DM). DR is usually identified through retinal image analysis, and in some countries, such as Brazil, such diagnosis is hindered by limited access to specialized care, leading to lengthy waits for ophthalmological evaluations. This scenario makes the Brazilian Diabetic Society’s annual DR check recommendation impractical for many. To address this gap, our study introduces a novel Siamese Convolutional Neural Network (SCNN) for DR prediction, usable by primary care professionals. Our SCNN analyzes pairs of eye fundus images and employs shared weights in its layers to extract essential features, facilitating similarity measurement between neural network outputs. Despite challenges with limited and imbalanced datasets, our SCNN showed effectiveness. We tested it on four datasets (IDRiD, APTOS, Messidor-1, DIARETDB0), with accuracy ranging from 67.23% (APTOS) to 96.85% (DIARETDB0). Compared to other methods, our approach consistently excelled, especially in recall analysis. These results suggest that deep learning via Siamese networks is likely to be a viable and potential DR screening tool.

WLAN monopole antenna design by Siamese convolutional neural network and KNN exploiting Gaussian process

In the process of antenna design, surrogate models can generally be used, but modeling requires a large number of samples. Although full wave electromagnetic simulation software can handle this task, obtaining a large number of samples is time-consuming, however too small number of sample may lead to lower accuracy of the trained surrogate model. Inspired by semi-supervised learning methods, this paper uses Siamese convolutional neural networks (SCNN) and K-nearest neighbor (KNN) algorithms to generate highly reliable virtual samples and expand the training sample set, further improving the accuracy and robustness of the surrogate model by exploiting Gaussian process (GP) models. The proposed method is named SCNN-KNN-GP, which is used for the design of WLAN dual band monopole antennas. Moreover, the relationships between the performance of the proposed model and the increased number of virtual samples and the coefficient of the KNN are studied, resulting in a more excellent surrogate model structure.

A Siamese Convolutional Neural Network for Identifying Mild Traumatic Brain Injury and Predicting Recovery.

Timely diagnosis of mild traumatic brain injury (mTBI) remains challenging due to the rapid recovery of acute symptoms and the absence of evidence of injury in static neuroimaging scans. Furthermore, while longitudinal tracking of mTBI is essential in understanding how the diseases progresses/regresses over time for enhancing personalized patient care, a standardized approach for this purpose is not yet available. Recent functional neuroimaging studies have provided evidence of brain function alterations following mTBI, suggesting mTBI-detection models can be built based on these changes. Most of these models, however, rely on manual feature engineering, but the optimal set of features for detecting mTBI may be unknown. Data-driven approaches, on the other hand, may uncover hidden relationships in an automated manner, making them suitable for the problem of mTBI detection. This paper presents a data-driven framework based on Siamese Convolutional Neural Network (SCNN) to detect mTBI and to monitor the recovery state from mTBI over time. The proposed framework is tested on the cortical images of Thy1-GCaMP6s mice, obtained via widefield calcium imaging, acquired in a longitudinal study. Results show that the proposed model achieves a classification accuracy of 96.5%. To track the state of the injured brain over time, a reference distance map is constructed, which together with the SCNN model, are employed to assess the recovery state in subsequent sessions after injury, revealing that the recovery progress varies among subjects. The promising results of this work suggest that a similar approach could be potentially applicable for monitoring recovery from mTBI, in humans.

Dimension-Expanded-Based Matching Method With Siamese Convolutional Neural Networks for Gravity-Aided Navigation

Matching algorithm is the key technique of the gravity-aided inertial navigation system. With the development of artificial intelligence, many neural network-based matching methods have been extensively studied. The pattern recognition-based matching methods transforms the matching problem as pattern recognition, which cannot be used directly on datasets where the neural networks have not been trained. To improve the accuracy of navigation and positioning, it is necessary to extract muti-dimensional gravity features from the limited navigation information. In this paper, the sequence of the gravity anomaly value is expanded to two-dimensional feature map containing time series features by Gramian Angular Fields method, which preserves the numerical information of the one-dimensional sequence and extracts the correlation relationship between each element. In addition, to reduce the influence of gravity measurement instrument error on the position precision of gravity matching algorithm, affine transformation is performed on INS trajectory and a Siamese convolutional neural network model is proposed to compare the measured gravity database with the gravity anomaly value in the pre-stored gravity background map and get the matching position. Simulation results and practical tests show that the proposed method can obtain a more precise location result compared with the traditional matching algorithm.

Few-shot transfer learning for personalized atrial fibrillation detection using patient-based siamese network with single-lead ECG records

The proliferation of wearable devices has allowed the collection of electrocardiogram (ECG) recordings daily to monitor heart rhythm and rate. For example, 24-hour Holter monitors, cardiac patches, and smartwatches are widely used for ECG gathering and application. An automatic atrial fibrillation (AF) detector is required for timely ECG interpretation. Deep learning models can accurately identify AFs if large amounts of annotated data are available for model training. However, it is impractical to request sufficient labels for ECG recordings for an individual patient to train a personalized model. We propose a Siamese-network-based approach for transfer learning to address this issue. A pre-trained Siamese convolutional neural network is created by comparing two labeled ECG segments from the same patient. We sampled 30-second ECG segments with a 50% overlapping window from the ECG recordings of patients in the MIT-BIH Atrial Fibrillation Database. Subsequently, we independently detected the occurrence of AF in each patient in the Long-Term AF Database. By fine-tuning the model with the 1, 3, 5, 7, 9, or 11 ECG segments ranging from 30 to 180 s, our method achieved macro-F1 scores of 96.84%, 96.91%, 96.97%, 97.02%, 97.05%, and 97.07%, respectively.

Despeckling of SAR Images Using Residual Twin CNN and Multi-Resolution Attention Mechanism

The despeckling of synthetic aperture radar images using two different convolutional neural network architectures is presented in this paper. The first method presents a novel Siamese convolutional neural network with a dilated convolutional network in each branch. Recently, attention mechanisms have been introduced to convolutional networks to better model and recognize features. Therefore, we propose a novel design for a convolutional neural network using an attention mechanism for an encoder–decoder-type network. The framework consists of a multiscale spatial attention network to improve the modeling of semantic information at different spatial levels and an additional attention mechanism to optimize feature propagation. Both proposed methods are different in design but they provide comparable despeckling results in subjective and objective measurements in terms of correlated speckle noise. The experimental results are evaluated on both synthetically generated speckled images and real SAR images. The methods proposed in this paper are able to despeckle SAR images and preserve SAR features.

A Siamese Inverted Residuals Network Image Steganalysis Scheme based on Deep Learning

With the rapid proliferation of urbanization, massive data in social networks are collected and aggregated in real time, making it possible for criminals to use images as a cover to spread secret information on the Internet. How to determine whether these images contain secret information is a huge challenge for multimedia computing security. The steganalysis method based on deep learning can effectively judge whether the pictures transmitted on the Internet in urban scenes contain secret information, which is of great significance to safeguarding national and social security. Image steganalysis based on deep learning has powerful learning ability and classification ability, and its detection accuracy of steganography images has surpassed that of traditional steganalysis based on manual feature extraction. In recent years, it has become a hot topic of the information hiding technology. However, the detection accuracy of existing deep learning based steganalysis methods still needs to be improved, especially when detecting arbitrary-size and multi-source images, their detection efficientness is easily affected by cover mismatch. In this manuscript, we propose a steganalysis method based on Inverse Residuals structured Siamese network (abbreviated as SiaIRNet method, Sia mese- I nverted- R esiduals- Net work Based method). The SiaIRNet method uses a siamese convolutional neural network (CNN) to obtain the residual features of subgraphs, including three stages of preprocessing, feature extraction, and classification. Firstly, a preprocessing layer with high-pass filters combined with depth-wise separable convolution is designed to more accurately capture the correlation of residuals between feature channels, which can help capture rich and effective residual features. Then, a feature extraction layer based on the Inverse Residuals structure is proposed, which improves the ability of the model to obtain residual features by expanding channels and reusing features. Finally, a fully connected layer is used to classify the cover image and the stego image features. Utilizing three general datasets, BossBase-1.01, BOWS2, and ALASKA#2, as cover images, a large number of experiments are conducted comparing with the state-of-the-art steganalysis methods. The experimental results show that compared with the classical SID method and the latest SiaStegNet method, the detection accuracy of the proposed method for 15 arbitrary-size images is improved by 15.96% and 5.86% on average, respectively, which verifies the higher detection accuracy and better adaptability of the proposed method to multi-source and arbitrary-size images in urban scenes.

Sensor fault analysis of aero-engine using ensemble SCNN and Bayesian interval estimation

Sensors of an aircraft engine have the problems of multiple kinds of faults and difficulty to diagnose. Moreover, the diagnosis performance cannot meet practical demands. Therefore, this paper proposes a novel approach for sensor fault diagnosis integrating the advantages of selective ensemble learning, siamese convolutional neural network (SCNN), and Bayesian interval estimation (BIE). The Bayesian interval estimation with hypothesis testing is developed with real-time data update to reduce the confidence interval. The attention mechanism is introduced to enhance fault diagnosis performance. Moreover, ensemble selection learning with Akaike information criteria is proposed to obtain optimal SCNNs. The theorems of convergence and optimal confidence data in ensemble learning are proven. Furthermore, this paper develops a novel improved learning algorithm based on AdaMod (Adaptive and momental bound method) and SGD (stochastic gradient descent), and achieves a new fault diagnosis algorithm to improve the diagnosis performance. The sensor data of aircraft engines from China Eastern Airlines is adopted for validation and experimental results show that proposed approach reaches an excellent tradeoff between sensor diagnosis performance, confidence interval, and runtime. The ablation studies are performed to analyze the role of leaner’s components. The proposed method improves training accuracy by 6.2%, fault diagnosis performance by 5.1%, fault prediction mean result by 4.6%, and fault prediction variance is reduced by 38.1% than the state-of-the-art methods.

Automated fundus ultrasound image classification based on siamese convolutional neural networks with multi-attention

Fundus ultrasound image classification is a critical issue in the medical field. Vitreous opacity (VO) and posterior vitreous detachment (PVD) are two common eye diseases, Now, the diagnosis of these two diseases mainly relies on manual identification by doctors. This method has the disadvantages of time-consuming and manual investment, so it is very meaningful to use computer technology to assist doctors in diagnosis. This paper is the first to apply the deep learning model to VO and PVD classification tasks. Convolutional neural network (CNN) is widely used in image classification. Traditional CNN requires a large amount of training data to prevent overfitting, and it is difficult to learn the differences between two kinds of images well. In this paper, we propose an end-to-end siamese convolutional neural network with multi-attention (SVK_MA) for automatic classification of VO and PVD fundus ultrasound images. SVK_MA is a siamese-structure network in which each branch is mainly composed of pretrained VGG16 embedded with multiple attention models. Each image first is normalized, then is sent to SVK_MA to extract features from the normalized images, and finally gets the classification result. Our approach has been validated on the dataset provided by the cooperative hospital. The experimental results show that our approach achieves the accuracy of 0.940, precision of 0.941, recall of 0.940, F1 of 0.939 which are respectively increased by 2.5%, 1.9%, 3.4% and 2.5% compared with the second highest model.

Fault Diagnosis of Wind Turbine with Alarms Based on Word Embedding and Siamese Convolutional Neural Network

Alarms generated by a wind turbine alarm system indicate the need for emergency action by operators to protect the turbine from running into risky conditions. However, it can be challenging for operators to identify the fault types that trigger alarms, particularly with few labeled fault samples. This paper proposes a novel fault diagnosis method for wind turbines with alarms that collaboratively uses labeled and unlabeled alarms to improve diagnosis accuracy. First, the proposed method distinguishes different alarm sequences using a designed Siamese convolutional neural network with an embedding layer (S-ECNN) model. Then, the fault category of an unknown alarm sequence is diagnosed based on similarity scores. Specifically, the Skip-gram model is used to mine potential relationships among alarms in unlabeled alarm sequences, and pretrained alarm vectors are obtained. In the S-ECNN model, the pretrained alarm vectors are further optimized and trained using labeled alarm sequences. The similarity scores are calculated based on the distance between the extracted discriminative features of alarm sequences. The effectiveness of the proposed method is validated using actual alarm data from a wind farm.

Using a lightweight Siamese neural network for generating a feature vector in a vascular authentication system

The article analyzes the possibility of using a Siamese convolutional neural network to solve the problem of vascular authentication on an embedded hardware platform with limited computing resources (Orange Pi One). The authors give a brief review of modern methods for calculating image feature vectors used in the tasks of classifying, comparing or searching for images by content: based on variational series (histograms), local descriptors, singular point descriptors, descriptors based on hash functions, neural network descriptors. They suggest using the architecture of a biometric authentication system (BAS) based on images of palms in the visible and near-IR spectra based on a Siamese convolutional neural network. The developed software solution allows using the Siamese neural network in the "full network" (both symmetrical channels of the neural network are used) and "half of the neural network" (only one channel is used) modes to reduce the time for comparing biometric data vectors - images of the palms of registered BAS users. The authors demonstrate advantages of the neural network features: universality, scalability and competitiveness, including on embedded hardware and software solutions with limited computing resources without graphics accelerators. The studies have shown that using the Siamese neural network, the "overall accuracy" of palm image classification can be improved from 0.929 to 0.968 when compared with the image vectorization method based on a perceptual hash, while showing a comparable authentication time for individuals registered in BAS. In the experiments, the authors use a database of 2,000 images for 400 people.