Deep Neural Network Framework Research Articles

Surveillance System for Real Time High Precision Recognition of Criminal

In an era marked by escalating concerns over public safety and crime prevention, the demand for sophisticated surveillance technologies has surged. This paper presents a groundbreaking solution: a Surveillance System for Real-Time High-Precision Recognition of Criminals. Leveraging advancements in deep learning and facial recognition, our system transcends the limitations of traditional surveillance setups by autonomously identifying individuals of interest in real-time. By employing a meticulously designed deep neural network framework, our system detects and tracks criminal faces amidst dynamic, crowded environments, effectively enhancing public safety measures. Through a seamless integration of cutting-edge technology and meticulous methodology, our innovative system heralds a new era in proactive crime prevention, offering unparalleled precision and efficiency in identifying and apprehending high-risk individuals. Keywords — Real-time surveillance, High-precision facial recognition, Crime prevention and Deep learning-based system

Unsteady flow-field forecasting leveraging a hybrid deep-learning architecture

With the rapid development of data-driven technology, deep learning has been widely used to study unsteady flow phenomena, owing to its high-dimensional, nonlinear, and excellent big-data processing capabilities. Rapidly and accurately predicting unsteady flow fields has been a major challenge in fluid mechanics. Therefore, we designed a new U-shaped hybrid deep neural network (UDNN) framework using a multilayer convolution. Through the multilayer learning interaction of high-dimensional flow-field data, the temporal and spatial characteristics of the flow-field flow are captured, and the flow-field characteristics are predicted in an end-to-end form. The UDNN comprises a convolutional neural layer, deconvolutional layer, convolutional long-term and short-term layers, and attention-mechanism layer. First, based on computational fluid dynamics, we generated unsteady flow-field datasets of the flow around fixed and rotating cylinders at different Reynolds numbers, which were used as training samples for the network framework. Second, we designed a U-shaped convolutional layer, added horizontal time-series feature processing and attention-mechanism units, and fused the deep feature information predicted by the model with shallow semantic information to predict the flow-field features. In addition, we compared the UDNN, proper orthogonal decomposition – long short-term memory, and traditional convolutional autoencoder – long short-term memory models in terms of the flow-field prediction error, model training time, and inference speed. The final results showed that the proposed UDNN framework achieved high accuracy and strong robustness in predicting unsteady flow fields.

Performance Prediction of Power Beacon-Aided Wireless Sensor-Powered Non-Orthogonal Multiple-Access Internet-of-Things Networks under Imperfect Channel State Information

In this paper, we investigate a novel power beacon (PB)-aided wireless sensor-powered non-orthogonal multiple-access (NOMA) Internet-of-Things (IoT) network under imperfect channel state information (CSI). Furthermore, the exact expression outage probability (OP) of two IoT users is derived to analyze the performance of the considered network. To give further insight, the expression asymptotic OP and diversity order are also expressed when the transmit power at the PB goes to infinity. Furthermore, a deep neural network (DNN) framework is proposed to concurrently forecast IoT users’ OP in relation to real-time setups for IoT users. Additionally, when compared to the traditional analysis, our created DNN shows the shortest run-time prediction, and the outcomes predicted by the DNN model almost match those of the simulation. In addition, numerical results validate our analysis, simulation, and prediction through a Monte Carlo Simulation. Furthermore, the results show the impact of the main parameter on our proposed system. Finally, these findings show that NOMA performs better than the conventional orthogonal multiple-access (OMA) techniques.

Disaggregating the Carbon Exchange of Degrading Permafrost Peatlands Using Bayesian Deep Learning

AbstractExtensive regions in the permafrost zone are projected to become climatically unsuitable to sustain permafrost peatlands over the next century, suggesting transformations in these landscapes that can leave large amounts of permafrost carbon vulnerable to post‐thaw decomposition. We present 3 years of eddy covariance measurements of CH4 and CO2 fluxes from the degrading permafrost peatland Iškoras in Northern Norway, which we disaggregate into separate fluxes of palsa, pond, and fen areas using information provided by the dynamic flux footprint in a novel ensemble‐based Bayesian deep neural network framework. The 3‐year mean CO2‐equivalent flux is estimated to be 106 gCO2 m−2 yr−1 for palsas, 1,780 gCO2 m−2 yr−1 for ponds, and −31 gCO2 m−2 yr−1 for fens, indicating that possible palsa degradation to thermokarst ponds would strengthen the local greenhouse gas forcing by a factor of about 17, while transformation into fens would slightly reduce the current local greenhouse gas forcing.

Multitask Learning Deep Neural Networks Enable Embedded Design of Active Metamaterials.

In this study, we propose and implement a deep neural network framework based on multitask learning aimed at simplifying the forward modeling and inverse design process of photonic devices integrating active metasurfaces. We demonstrate and validate our approach by constructing a continuously tunable bandpass filter that is effective in the midwave infrared region. The key to this filter is the combination of a metasurface and Fabry-Perot (F-P) cavity structure of the tunable phase-change material Ge2Sb2Se4Te (GSST) and the precise control of the crystallinity of the GSST by a silicon-based heater. With the help of a deep learning framework, we are able to independently model the crystallinity and geometric parameters of the filter to maximize the use of GSST tuning for bandpass filtering. Our model discusses the self-attention mechanism and the effect of noise and compares several existing popular algorithms, and the results show that a multitask deep learning strategy can better assist the on-demand reverse design of photonic structures with phase change materials. This opens up new possibilities for personalization and functional extension of optical devices.

Deep Adaptive Fuzzy Clustering for Evolutionary Unsupervised Representation Learning.

Cluster assignment of large and complex datasets is a crucial but challenging task in pattern recognition and computer vision. In this study, we explore the possibility of employing fuzzy clustering in a deep neural network framework. Thus, we present a novel evolutionary unsupervised learning representation model with iterative optimization. It implements the deep adaptive fuzzy clustering (DAFC) strategy that learns a convolutional neural network classifier from given only unlabeled data samples. DAFC consists of a deep feature quality-verifying model and a fuzzy clustering model, where deep feature representation learning loss function and embedded fuzzy clustering with the weighted adaptive entropy is implemented. We joint fuzzy clustering to the deep reconstruction model, in which fuzzy membership is utilized to represent a clear structure of deep cluster assignments and jointly optimize for the deep representation learning and clustering. Also, the joint model evaluates current clustering performance by inspecting whether the resampled data from estimated bottleneck space have consistent clustering properties to improve the deep clustering model progressively. Experiments on various datasets show that the proposed method obtains a substantially better performance for both reconstruction and clustering quality compared to the other state-of-the-art deep clustering methods, as demonstrated with the in-depth analysis in the extensive experiments.

Explainable Artificial Intelligence Approach for Diagnosing Faults in an Induction Furnace

For over a century, induction furnaces have been used in the core of foundries for metal melting and heating. They provide high melting/heating rates with optimal efficiency. The occurrence of faults not only imposes safety risks but also reduces productivity due to unscheduled shutdowns. The problem of diagnosing faults in induction furnaces has not yet been studied, and this work is the first to propose a data-driven framework for diagnosing faults in this application. This paper presents a deep neural network framework for diagnosing electrical faults by measuring real-time electrical parameters at the supply side. Experimental and sensory measurements are collected from multiple energy analyzer devices installed in the foundry. Next, a semi-supervised learning approach, known as the local outlier factor, has been used to discriminate normal and faulty samples from each other and label the data samples. Then, a deep neural network is trained with the collected labeled samples. The performance of the developed model is compared with several state-of-the-art techniques in terms of various performance metrics. The results demonstrate the superior performance of the selected deep neural network model over other classifiers, with an average F-measure of 0.9187. Due to the black box nature of the constructed neural network, the model predictions are interpreted by Shapley additive explanations and local interpretable model-agnostic explanations. The interpretability analysis reveals that classified faults are closely linked to variations in odd voltage/current harmonics of order 3, 11, 13, and 17, highlighting the critical impact of these parameters on the model’s prediction.

CriticalFuzz: A critical neuron coverage-guided fuzz testing framework for deep neural networks

Context:Deep neural networks (DNN) have been widely deployed in safety-critical domains, such as autonomous cars and healthcare, where error behaviors can lead to serious accidents, testing DNN is extremely important. Neuron coverage-guided fuzz testing (NCFT) has become an effective whitebox testing approach for testing DNN, which iteratively generates new test cases with the guidance of neuron coverage to explore different logics of DNN, and has found numerous defects. However, existing NCFT approaches ignore that the role of neurons is distinct for the final output of DNN. Given an input, only a fraction of neurons determines the final output of the DNN. These neurons hold the essential logic of the DNN. Objective:To ensure the quality of DNN and improve testing efficiency, NCFT should first cover neurons containing major logic of DNN. Method:In this paper, we propose the critical neurons that hold essential logic of DNN. In order to prioritize the detection of potential defects of critical neurons, we propose a fuzz testing framework, named CriticalFuzz, which mainly contains the energy-based test case generation and the critical neuron coverage criteria. The energy-based test case generation has the capability to produce test cases that are more likely to cover critical neurons and involves energy-based seed selection, power schedule, and seed mutation. The critical neuron coverage as a mechanism for providing feedback to guide the CriticalFuzz in prioritizing the coverage of critical neurons. To evaluate the significance of critical neurons and the performance of CriticalFuzz, we conducted experiments on popular DNNs and datasets. Results:The experiment results show that (1) the critical neurons have a 100% impact on the output of models, while the non-critical neurons have a lesser effect; (2) CriticalFuzz is effective in achieving 100% coverage of critical neurons and covering 10 classes of critical neurons, outperforming both DeepHunter and TensorFuzz. (3) CriticalFuzz exhibits exceptional error detection capabilities, successfully identifying thousands of errors across 10 diverse error classes within DNN. Conclusion:The critical neurons defined in this paper hold more significant logic of DNN than non-critical neurons. CriticalFuzz can preferentially cover critical neurons, thereby improving the efficiency of the NCFT process. Additionally, CriticalFuzz is capable of identifying a greater number of errors, thus enhancing the reliability and effectiveness of the NCFT.

Recurrent neural network (RNN) and long short-term memory neural network (LSTM) based data-driven methods for identifying cohesive zone law parameters of nickel-modified carbon nanotube reinforced sintered nano-silver adhesives

In this paper, we presented three neural network models including deep neural network (DNN), recurrent neural network (RNN), and long short-term memory neural network (LSTM), which are proposed to predict the cohesive zone parameter of sintered silver DCB joint with different contents of nickel modified carbon nanotube, thus reducing the complexity of CZM parameter acquisition for nanoparticle reinforced adhesive. The bilinear CZM model is used as the prediction target model for sintered silver joints with different contents of nickel-modified carbon nanotube filler, and data sets suitable for different networks are established through experimental and numerical simulation results. Three kinds of networks are trained based on the optimized hyperparameters obtained from the Bayesian hyperparameters tuning process. The results show that DNN, RNN, and LSTM frameworks can all predict CZM parameters of nanoparticle-reinforced sintered silver adhesive through load-displacement curves. Based on loss analysis and statistical indicator comparison after K-fold cross-validation, the RNN and LSTM models have better prediction accuracy and performance than the DNN model, and the accuracy of the LSTM model is further improved compared with the DNN model. RNN and LSTM models have high prediction accuracy and stronger recognition ability for the time series data, they can be used as suitable alternative models for inverse recognition of CZM parameters of nanoparticle-reinforced adhesives, and have broad application prospects.

Deep transfer learning strategy in intelligent fault diagnosis of gas turbines based on the Koopman operator

The gas turbine engine is a predominant prime mover in the transport and energy sectors, and ensuring its reliable operation holds paramount significance. While intelligent fault diagnosis (FD) approaches have seen successful advancements within the gas turbine FD landscape, many existing methods operate under the assumption of identical health states during both data collection and the FD process. Moreover, most previous studies have overlooked the diagnosis of both sensors and actuators. Another critical challenge lies in isolating simultaneous and multiple faults and providing compromising FD performance, especially in the face of continued system performance degradation. Aiming at these problems, this study develops a novel unsupervised data-driven FD strategy based on leveraging the potential of transfer learning and the Koopman operator. A deep neural network-based transfer learning framework is proposed for realizing a precise adaptive linear model called the deep transfer linear (DTL) model enabling reliable prediction of the system’s behavior in various situations and designing structured fault residuals. To this end, a deep neural network framework is used to obtain a precise Koopman model realization using the richly collected data in the source domain. Subsequently, the realized model is fine-tuned for the target domains associated with the degraded system, mitigating the adverse effects of domain shift and addressing the rich data scarcity problem in the target domain. In addition, the dedicated and generalized residual sets are designed and generated employing the geometric approach for fault isolation and a decision-making analysis is developed to diagnose simultaneous faults. The reliability of the proposed strategy is demonstrated through various experiments in the presence of noise and performance degradation, and a comparative performance analysis is conducted between the proposed strategy and another data-driven method showcasing the superiority of the proposed approach.

Geometrically-informed predictive modeling of melt pool depth in laser powder bed fusion using deep MLP-CNN and metadata integration

The laser powder bed fusion (LPBF) process has the ability to manufacture intricate geometries. However, the fabrication of complex geometries using the LPBF process for industrial applications remains challenging in terms of achieving consistent microstructure and mechanical properties. Experimental investigations show that in addition to process parameters, geometrical factors such as shape and size of parts impact the thermal history, microstructure, defect structure, and thus mechanical and fatigue properties. As a result, the standardization, qualification, and certification of additively manufactured (AM-ed) parts encounters substantial obstacles posed by the obscure impact of design parameters on component quality. In this paper, an artificial intelligence framework is proposed to predict the melt pool depth of the additively manufactured parts considering the geometrical information of the parts. More specifically, a deep Multilayer Perceptron and Convolutional Neural Network (MLP-CNN) framework is designed to predict melt pool depth for any part geometries when printing metallic parts via LPBF using metadata (i.e., numerical and image data). Quite often the melt pool depth generated during the AM processes is employed as a surrogate for thermal, defect- and micro-structural signatures. In melt pool depth prediction of complex geometries in LPBF, a digital twin environment using a finite element model (FEM) is developed and validated using experimental melt pool measurements for different geometries within multi-tracks and layers. The FEM was in agreement with experimentation with an error of less than 15%. Through the process simulation, a large dataset of melt pool depths was obtained for various part geometries and different process parameters. Next, the MLP-CNN framework was established to identify the parallel impact of part design and process parameters on melt pool depths. To enhance the performance and robustness of this model, data augmentation has been implemented by rotating and transferring geometries to artificially expand the dataset. Data augmentation helps to mitigate overfitting and promote better generalization, especially in the context of limited training data. After training, the proposed model is found to give accurate melt pool prediction even for new geometries not considered during training. The fusion of CNN and MLP has led the melt pool predictions for unseen geometries with an accuracy of 95%.

FLEX: A fast and light-weight learned index for kNN search in high-dimensional space

The k Nearest Neighbors (kNN) search in high-dimensional space is a fundamental problem with various applications. In this paper, we try to solve this problem by using deep neural networks (DNNs). We apply DNNs to represent complex correlations between high-dimensional objects, enabling us to project similar objects into the same class, thus reducing the search cost. Based on DNNs, we propose two novel techniques to improve query efficiency while keeping accuracy. First, traditional DNNs typically demand extensive training data for achieving high accuracy. To decrease the training size, we design a multi-module DNN framework comprising several small modules. Each module learns to capture part of knowledge for the given query. The collective output of these sub-modules is then seamlessly integrated to form the final result. Second, with machine learning models, the size of candidates are unbounded. Thus, we design a linear-time data layout refinement algorithm, aiming to limit the number of candidates to a small constant. Empirically we find that our approach significantly outperforms the state-of-the-art methods in terms of both time efficiency and space efficiency while still attaining comparable or better accuracy.

A deep learning framework for prediction of crop yield in Australia under the impact of climate change

Accurate prediction of crop yields is essential to ensure food security. In this study, a new deep neural networks framework is developed to predict crop yields in Australia, considering the impact of climate change, fertilizer use, and crop area. It is implemented for oats, corn, rice, and wheat crops, and its forecasting performance is benchmarked against five statistical and machine learning methods. All the software codes for the implementation of the proposed framework are freely available. The proposed framework shows the highest forecasting performance for all the considered crop types. It provides 23%, 38%, 39%, and 40% lower average mean absolute error than the benchmark methods for oat, corn, rice, and wheat crops, respectively. The reductions in average root mean squared error are 19%, 25%, 37%, and 29% over the benchmark methods. Then, it is used to predict yields of the considered crops in Australia towards 2025 under six different climate change scenarios. It is observed that although climate change has some boosting impact on crop yield, it is not sustainable to meet the demand. However, it is possible to keep crop yields rising while mitigating climate change.

DeepAUP: A Deep Neural Network Framework for Abnormal Underground Heat Transport Pipelines

DeepAUP: A Deep Neural Network Framework for Abnormal Underground Heat Transport Pipelines

An efficient cloud-integrated distributed deep neural network framework for IoT malware classification

The proliferation of interconnected devices in the Internet of Things (IoT) landscape has introduced significant security concerns. With the integration of android devices, the potential for attackers to exploit vulnerabilities becomes a crucial issue. Timely detection of malware attacks has emerged as a critical challenge, especially in industrial IoT (IIoT), to prevent permanent damage. The slow inference speed of the current algorithms causes limited performance and a significant delay in the detection and mitigation of malware, especially in large-scale environments such as IIoT. High resource utilization can limit the size of the model (number of parameters), which could be implemented in embedded devices, which causes lower accuracy of the model. Hence, we proposed a HierarchicalCloudDNN, a distributed malware detection framework that used modified state-of-the-art deep learning algorithms in a boosting ensemble architecture. The proposed method could scale efficiently from IIoT devices up to the edge and cloud to speed up malware detection, simultaneously reduce resource utilization, and maintain a controllable level of accuracy compared to rival methods in malware detection. The objective is to minimize both run-time and resource consumption while maintaining performance characteristics. The performance analysis of the HierarchicalCloudDNN method was evaluated using standard performance measures on the frequently used BIG 2015 dataset. Based on evaluations, the proposed approach has demonstrated an impressive accuracy rate of 98.90%, outperforming many state-of-the-art models in performance, response time, and resource utilization. These findings suggest a notable decrease in both false positives and false negatives while maintaining accuracy. The proposed distributed malware analysis framework is scalable, robust, and helpful for the timely detection of malicious activity, especially in geographically distributed environments such as the IIoT.

Application of generative adversarial networks in image, face reconstruction and medical imaging: challenges and the current progress

ABSTRACT In deep learning, GANs (Generative Adversarial Networks) are one of the prominent study areas due to their ability to generate synthetic data thereby solving the problem of the unavailability and the limited data sets. GAN is a framework of deep neural networks that can learn from a set of training data and generate new data with similar characteristics as the training data. In this review, the history of GAN, various types of GANs, objective functions, loss functions, and performance analysis done for GANs in various fields are also analysed. The main objective of the paper is to analyse the application of GANs in face restoration, and medical imaging including their evaluation metrics and data sets used. A deep review has been carried out on various types of GANs, their architecture, objective functions, and applications. This review focuses more on the medical applications using various types of GANs including image augmentation, disease detection, medical image enhancement, face restoration, detection, etc. The challenges, current progress, and future applications using various GANs are also discussed. This review clearly shows that the application of GANs has increased considerably and thus it proves a promising future in the field of deep learning.

MLCapsNet +: A multi-capsule network for the identification of the HIV ISs along important sequence positions

The most studied sub-category of the retrovirus is the human immunodeficiency virus (HIV), which is a type of virus of the Retroviridae family. The HIV integration site (HIV IS)/ integration sites (HIV ISs) denote a crucial entity in the entire process of infection and its rebound if there is an interruption in therapy. It determines the steps involved in the formation of latent viral reserve. This work proposes a very deep neural network framework, where each of the layers of the multi-layered network is itself a neural network (NN). The attention mechanism is used for the extraction of the importance of positions in terms of an attention map. This framework will use the Capsule Network along with the attention mechanism to increase the explainability about the presence of local features. Convolutional neural networks (CNNs), which are specialized for image-based recognition and classification, have many drawbacks with one of these being its invariance to translation. This drawback has been overcome by the Capsule Networks. The proposed model also identifies the HIV ISs achieving a performance better than the State-of-the-art methods. Support Vector Machine (SVM), Random Forest (RF) and Logistic Regression (LR) classifiers have been used. Only two state-of-the-art methods are present in the literature that achieve these two goals and a comparison of this work with those two works has been provided here. This work performs way better than the state-of-the-art work in detecting the HIV IS. The comparison is presented here in terms of AUC-ROC, AUC-PR, accuracy, confusion matrix, and Fβ score.

New approach of milling state recognition based on attention mechanism weighted fusion K-nearest neighbor under framework of chaotic attractor using accelerated life signal

The surface quality of the finished workpiece and the lifetime of the machining tool will be damaged and reduced due to the issue of untimely milling monitoring and identification. To alleviate this bottleneck, in this paper, a new milling state recognition approach named hybrid attention mechanism WFKNN model under the framework of chaotic attractor and deep convolutional neural network (CNN) is proposed during the milling process. Specifically, three milling stages, that is, early-term stage, mid-term stage, and ultimate-term milling stage, are divided through Duffing chaotic attractor (DCA) characteristic, where the periodicity and vorticity phenomena under different milling states will be revealed using DCA characteristic. Then, the signal features with respect to the milling condition are extracted by the CNN-attention mechanism strategy, and the extracted signal features will be fed into the improved KNN classifier optimized by improving the distance metric. Eventually, the availability and feasibility of the proposed approach is verified via six group run-to-failure datasets of the milling process. The results show that a higher identify accuracy is obtained via the proposed approach compared with existing benchmarks such as the CNN model.

Deep convolutional neural network framework with multi-modal fusion for Alzheimer’s detection

The biomedical profession has gained importance due to the rapid and accurate diagnosis of clinical patients using computer-aided diagnosis (CAD) tools. The diagnosis and treatment of Alzheimer’s disease (AD) using complementary multimodalities can improve the quality of life and mental state of patients. In this study, we integrated a lightweight custom convolutional neural network (CNN) model and nature-inspired optimization techniques to enhance the performance, robustness, and stability of progress detection in AD. A multi-modal fusion database approach was implemented, including positron emission tomography (PET) and magnetic resonance imaging (MRI) datasets, to create a fused database. We compared the performance of custom and pre-trained deep learning models with and without optimization and found that employing natureinspired algorithms like the particle swarm optimization algorithm (PSO) algorithm significantly improved system performance. The proposed methodology, which includes a fused multimodality database and optimization strategy, improved performance metrics such as training, validation, test accuracy, precision, and recall. Furthermore, PSO was found to improve the performance of pre-trained models by 3-5% and custom models by up to 22%. Combining different medical imaging modalities improved the overall model performance by 2-5%. In conclusion, a customized lightweight CNN model and nature-inspired optimization techniques can significantly enhance progress detection, leading to better biomedical research and patient care.

A single-stage anchor-free rotating target visual detection algorithm suitable for flexible body vibration displacement measurement

The phenomenon of angular inclination of flexible structures during vibration poses a significant challenge to the applicability of visual vibration measurement methods because the target locked in the captured image will produce unknown geometric deformations such as scale, displacement, and angle in the time domain space, and the horizontal rectangular frame used for matching during target detection will also increase the false detection rate of the target due to the introduction of more background information. Such subtle geometric deformations and false detections can lead to severe fit errors in the displacement curves regressed by the visual vibration measurement algorithm. To effectively improve the accuracy and robustness of vibration image target recognition, this article takes the flexible body captured by a high-speed camera as the target of vibration displacement measurement. It introduces the rotating target detection method based on deep learning into the field of visual vibration measurement, which verifies the feasibility of the deep learning method in flexible body vibration measurement, and based on the deep convolutional neural network framework, a high-precision displacement measurement algorithm based on single-stage anchor-free rotating target detection is proposed. The algorithm in this article first uses the CSPDarknet backbone network to extract multi-scale features of flexible structural image sequences. It then uses PANet to fuse the top-down and bottom-up bidirectional feature maps of the four bridge target feature maps obtained through the backbone network. The shallow and deep information is used for semantic feature fusion and combined with the Coordinate Attention mechanism to achieve target finding and fine positioning on the feature map. Finally, we use the coordinates of the bounding box obtained from the test to regress the position offset of the object’s center point. To verify the accuracy of the algorithm in this article, we conducted experimental validation on the cable-stayed bridge model and the actual bridge and compared the performance with the traditional template matching algorithm, differential optical flow method, and various deep learning algorithms with different localization principles, as well as the displacement signals collected and processed by accelerometers. The experimental results of time-frequency characteristics analysis show that the vibration displacement trajectories regressed by the algorithm in this paper have the best overlap with the displacement measurements collected by the accelerometer, which verifies that the algorithm in this article has good application potential and implementation space in the field of condition monitoring of flexible structural bodies.