Improve Identification Accuracy Research Articles

Implementing Deep Convolutional Neural Networks for QR Code-Based Printed Source Identification

QR codes (short for Quick Response codes) were originally developed for use in the automotive industry to track factory inventories and logistics, but their popularity has expanded significantly in the past few years due to the widespread applications of smartphones and mobile phone cameras. QR codes can be used for a variety of purposes, including tracking inventory, advertising, electronic ticketing, and mobile payments. Although they are convenient and widely used to store and share information, their accessibility also means they might be forged easily. Digital forensics can be used to recognize direct links of printed documents, including QR codes, which is important for the investigation of forged documents and the prosecution of forgers. The process involves using optical mechanisms to identify the relationship between source printers and the duplicates. Techniques regarding computer vision and machine learning, such as convolutional neural networks (CNNs), can be implemented to study and summarize statistical features in order to improve identification accuracy. This study implemented AlexNet, DenseNet201, GoogleNet, MobileNetv2, ResNet, VGG16, and other Pretrained CNN models for evaluating their abilities to predict the source printer of QR codes with a high level of accuracy. Among them, the customized CNN model demonstrated better results in identifying printed sources of grayscale and color QR codes with less computational power and training time.

Super-resolution compressive spherical beamforming based on off-grid sparse Bayesian inference

Compressive spherical beamforming (CSB) with spherical microphone arrays is a panoramic acoustic source identification technology with high spatial resolution and clear acoustic imaging, which has broad application prospects. Due to the discretization of the focus region and the assumption of on-grid sources, classical CSB suffers from the basis mismatch problem, i.e., it faces performance deterioration when identifying off-grid sources. To overcome the problem, this paper proposes off-grid sparse Bayesian inference-based CSB (OGSBI-CSB). OGSBI-CSB formulates the direction of arrival (DOA) as the sum of the DOA of grid point and DOA offset, constructs an off-grid model based on Taylor expansion, and adapts OGSBI to solve the model and obtains DOA and strength estimation. Simulations and experiments demonstrate that the proposed OGSBI-CSB not only can effectively alleviate the basis mismatch problem and then improve identification accuracy for off-grid sources, but also enjoys super-resolution and good resistance to noise interference.

3D ground penetrating radar cavity identification algorithm for urban roads using transfer learning

3D ground penetrating radar (GPR) is the main method for the detection of underground cavities in urban roads. The number of road cavity samples detected by 3D radar is small, whereas the intelligent identification model require a large number of learning samples for model training, resulting in inadequate model training. This causes the model to be less accurate in identifying cavities, resulting in many misses and misjudgments. Given the above problems, combined with the detection characteristics of the vertical, the horizontal, and the crossed slices obtained in one detection process of 3D GPR, a 3D GPR cavity intelligent recognition model based on model-based transfer learning is proposed. Firstly, a large amount of 3D GPR data of urban road models with cavities are obtained through forwarding simulation. And the intelligent recognition model was pre-trained on the cavity detection data on three types of slices respectively. Then, through model-based transfer learning, a small amount of real underground cavity data is used to speed up the convergence speed of model training and optimize the structural parameters. It breaks through the limitation of the insufficient number of cavity samples for 3D radar detection on the intelligent learning model training, reduces algorithm training costs, and improves identification accuracy.

Research on characteristic parameter selection and attention-GRU-based model for braking intention identification

The braking intention is of great significance to the realization of driver assistant features, the improvement of braking safety, and the maximization of energy recovery efficiency for electric vehicles. With the aim of accurate identification of braking intention, an identification model based on Gated Recurrent Unit (GRU) Network with Attention mechanism is proposed in this paper. Based on numerous vehicle braking test data, braking process analysis, characteristic parameters selection, identification model training, and verification are carried out. Through the difference analysis based on the Kruskal-Wallis test and the importance evaluation based on random forest, combined with the real-time requirements of practical application, the appropriate characteristic parameters are selected as the model input. The attention mechanism is introduced into the proposed model, which can improve identification accuracy by capturing valuable feature information. The comparative verification results show that the Attention-GRU model performs better than the other three comparison models, and its identification accuracy is 96.7%, of which the accuracy of slight braking, normal braking, and emergency braking are 96.3%, 95.8%, and 100% respectively. The identified braking intention can provide an effective basis for the establishment of vehicle control strategies.

Multiple Objects Automatic Detection of GPR Data Based on the AC-EWV and Genetic Algorithm

The automatic detection of multiple objects in ground penetrating radar (GPR) data is investigated by searching for the reflected hyperbolas of buried objects, which can reduce the subjectivity of operators and improve identification accuracy. Based on Frequency-wavenumber (F-K) migration, the accurate calculation of electromagnetic wave velocity (AC-EWV) is proposed by searching for the minimum image entropy of migrated radargrams. To avoid global searching, potential positions of object hyperbolas are selected from the binarized radargram through the vertical gray gradient searching, then the sub_window is extracted with the potential position as the center. The best fitting hyperbola is detected with the genetic algorithm (GA) in the sub_window, and objects are finally determined with five hyperbolic matching criteria and the auto-categorization. This technique is verified with the simulated and measured GPR data about rebars, pipelines, and voids, and results demonstrate that it achieves the average correct rate, average missed rate, and the average misjudged rate is 98.46%, 1.33%, and 0.36%, respectively, and the average correct rate for GPR data of the double-layer rebars is 91.67%.

Transformer Based Defense GAN Against Palm-Vein Adversarial Attacks

Vein biometrics is a high security and privacy preserving identification technology that has attracted increasing attention over the last decade. Deep neural networks (DNNs), such as convolutional neural networks (CNN), have shown strong capabilities for robust feature representation, and have achieved, as a result, state-of-the-art performance on various vision tasks. Inspired by their success, deep learning models have been widely investigated for vein recognition and have shown significant improvement of identification accuracy compared to handcrafted models. Existing deep learning models, however, are vulnerable to adversarial perturbation attacks, where thoughtfully crafted small perturbations can cause misclassification of legitimate images, degrading, thereby, the efficiency of vein recognition systems. To address this problem, we propose, in this paper, VeinGuard, a novel defense framework to defend deep learning classifiers against adversarial palm-vein image attacks, composed of a local transformer-based GAN and a purifier. VeinGuard comprises two components: a local transformer-based GAN (LTGAN) that learns the distribution of unperturbed vein images and generates high-quality palm-vein images, and a purifier consisting of a trainable residual network and of a pre-trained generator from LTGAN that automatically removes a wide variety of adversarial perturbations. The resulting clean images are fed to vein classifiers for identification, thereby avoiding adversarial attacks. We evaluate VeinGuard on three public vein datasets in terms of white-box attacks, black-box attacks, ablation experiments, and computation time. The experimental results show that VeinGuard allows filtering the perturbations and enables the classifiers to achieve state-of-the-art recognition results for different adversarial attacks.

Detection of Coconut Clusters Based on Occlusion Condition Using Attention-Guided Faster R-CNN for Robotic Harvesting.

Manual harvesting of coconuts is a highly risky and skill-demanding operation, and the population of people involved in coconut tree climbing has been steadily decreasing. Hence, with the evolution of tree-climbing robots and robotic end-effectors, the development of autonomous coconut harvesters with the help of machine vision technologies is of great interest to farmers. However, coconuts are very hard and experience high occlusions on the tree. Hence, accurate detection of coconut clusters based on their occlusion condition is necessary to plan the motion of the robotic end-effector. This study proposes a deep learning-based object detection Faster Regional-Convolutional Neural Network (Faster R-CNN) model to detect coconut clusters as non-occluded and leaf-occluded bunches. To improve identification accuracy, an attention mechanism was introduced into the Faster R-CNN model. The image dataset was acquired from a commercial coconut plantation during daylight under natural lighting conditions using a handheld digital single-lens reflex camera. The proposed model was trained, validated, and tested on 900 manually acquired and augmented images of tree crowns under different illumination conditions, backgrounds, and coconut varieties. On the test dataset, the overall mean average precision (mAP) and weighted mean intersection over union (wmIoU) attained by the model were 0.886 and 0.827, respectively, with average precision for detecting non-occluded and leaf-occluded coconut clusters as 0.912 and 0.883, respectively. The encouraging results provide the base to develop a complete vision system to determine the harvesting strategy and locate the cutting position on the coconut cluster.

Few-Shot Radar Emitter Signal Recognition Based on Attention-Balanced Prototypical Network

In recent years, radar emitter signal identification has been greatly developed via the utilization of deep learning and has achieved significant improvements in identification accuracy. However, with the continuous emergence of complex regime radars and the increasing complexity of the electromagnetic environment, some new kinds of radar emitter signals collected are not in sufficient quantities to satisfy the demand of deep learning. As a result, in this paper, we adopted the prototypical network (PN) belonging to metric-based meta-learning to realize few-shot radar emitter signal recognition with the aim of meeting the needs of modern electronic warfare. Additionally, considering the problems that may arise in the field of few-shot radar emitter signal recognition, such as discriminative location bias caused by a small number of base classes or the large difference between base classes and novel classes, we proposed an attention-balanced strategy to improve meta-learning. Specifically, each channel in the feature map is forced to make the same contribution in the distinguishment of different classes. In addition, for PN, taking into account that the feature vectors of each support sample in the class are different, we set a network to exploit the relation between each support sample in the same classes, and weighted each feature vector of the support samples according to the relation. Large quantities of experiments indicate that our algorithm possesses more advantages than other algorithms.

SwipePass

Pattern lock-based authentication has been widely adopted in modern smartphones. However, this scheme relies essentially on passwords, making it vulnerable to various side-channel attacks such as the smudge attack and the shoulder-surfing attack. In this paper, we propose a second-factor authentication system named SwipePass, which authenticates a smartphone user by examining the distinct physiological and behavioral characteristics embedded in the user's pattern lock process. By emitting and receiving modulated audio using the built-in modules of the smartphone, SwipePass can sense the entire unlocking process and extract discriminative features to authenticate the user from the signal variations associated with hand dynamics. Moreover, to alleviate the burden of data collection in the user enrollment phase, we conduct an in-depth analysis of users' behaviors under different conditions and propose two augmentation techniques to significantly improve identification accuracy even when only a few training samples are available. Finally, we design a robust authentication model based on CNN-LSTM and One-Class SVM for user identification and spoofer detection. We implement SwipePass on three off-the-shelf smartphones and conduct extensive evaluations in different real-world scenarios. Experiments involving 36 participants show that SwipePass achieves an average identification accuracy of 96.8% while maintaining a false accept rate below 0.45% against various attacks.

Identifying Protein Complexes From Protein-Protein Interaction Networks Based on Fuzzy Clustering and GO Semantic Information.

Protein complexes are of great significance to provide valuable insights into the mechanisms of biological processes of proteins. A variety of computational algorithms have thus been proposed to identify protein complexes in a protein-protein interaction network. However, few of them can perform their tasks by taking into account both network topology and protein attribute information in a unified fuzzy-based clustering framework. Since proteins in the same complex are similar in terms of their attribute information and the consideration of fuzzy clustering can also make it possible for us to identify overlapping complexes, we target to propose such a novel fuzzy-based clustering framework, namely FCAN-PCI, for an improved identification accuracy. To do so, the semantic similarity between the attribute information of proteins is calculated and we then integrate it into a well-established fuzzy clustering model together with the network topology. After that, a momentum method is adopted to accelerate the clustering procedure. FCAN-PCI finally applies a heuristical search strategy to identify overlapping protein complexes. A series of extensive experiments have been conducted to evaluate the performance of FCAN-PCI by comparing it with state-of-the-art identification algorithms and the results demonstrate the promising performance of FCAN-PCI.

Automated fatigue crack detection in steel box girder of bridges based on ensemble deep neural network

This paper constructs a novel ensemble deep neural network to improve identification accuracy of bridge cracks in complex backgrounds. In the three sub-networks of the ensemble network, the detection classifier distinguishes cracks and distractors at the scale of image patches, and then the segmentation sub-network obtains pixel-level crack details. Meanwhile, an innovative crack inference sub-network is constructed based on the prior knowledge of crack morphology. The inference sub-network can predict the probability of cracks existence using the learned image correlation between adjacent regions, and feedback to the detection and segmentation sub-networks. Then, the prediction probabilities coming from the image feature and morphological reasoning are fused to judge and correct the crack repeatedly. Therefore, the presented ensemble deep neural network can well simulate human multi-scale observation, reasoning and decision-making processes. The identification results show that the proposed algorithm can effectively reduce the misjudgment and provide more accurate crack segmentation.

A model fusion strategy for identifying aircraft risk using CNN and Att-BiLSTM

The Aviation Safety Reporting System (ASRS) data is an important information source for risk identification in civil aviation. However, ASRS data has the characteristics of high dimensionality, unstructured and data imbalance. This brings great challenges for risk identification. In this paper, a model fusion strategy is proposed for identifying the aircraft risk by using convolutional neural network (CNN) and bidirectional long short-term memory neural network with attention mechanism (Att-BiLSTM). Firstly, all obtained ASRS data are reasonably divided into five risk levels. Secondly, the CNN is utilized to deeply mine the relationship between risk levels and structured data. Meanwhile, the Att-BiLSTM is adopted to deeply mine the relationship between risk levels and unstructured event narratives. Finally, a model fusion strategy is proposed to fuse the preliminary identifications of CNN and Att-BiLSTM to obtain the final identifications. The adopted CNN and Att-BiLSTM are compared with two individual models respectively. The experimental results show that the CNN and Att-BiLSTM have superior identification performance. Furthermore, by comparing the performance of the proposed model fusion strategy against the effect of using the individual model, the proposed model fusion strategy is effective in improving identification accuracy and reducing the influence of data imbalance in civil aviation.

A multitask joint framework for real-time person search

Person searches generally involve three important parts: person detection, feature extraction and identity comparison. However, a person search integrating detection, extraction and comparison has the two following drawbacks. First, the accuracy of detection will affect the accuracy of comparison. Second, it is difficult to achieve real-time results in real-world applications. To solve these problems, we propose a multitask joint framework for real-time person search (MJF) that optimizes person detection, feature extraction and identity comparison. For the person detection module, we propose the YOLOv5-GS model, which is trained with a person dataset. YOLOv5-GS combines the advantages of the Ghostnet and the squeeze-and-excitation block and improves the speed of person detection. For the feature extraction module, we design a model adaptation architecture, which can select different networks according to the number of people. It can balance the relationship between accuracy and speed. For identity comparison, we propose a 3D pooled table and a matching strategy to improve identification accuracy. On the condition of 1920 \(\times\) 1080-resolution video and a 200-ID table, the IR and the FPS achieved by our method reach 82.69% and 25.14, respectively. Therefore, the MJF can achieve real-time person search.

Transformer helps identify kiwifruit diseases in complex natural environments

The complex background of disease images and the small contrast between the disease area and the background easily confuse, seriously affecting the robustness and accuracy of kiwifruit disease identification models. To address the above problems, this paper proposes a disease identification model based on Vision Transformer and Convolutional Neural Network, ConvViT(Convolutional Neural Network and Vision Transformer), to identify diseases by extracting effective features of kiwifruit disease spots. The proposed ConvViT includes convolutional structure and Transformer structure: The convolutional structure is used to extract the global features of the image, and the Transformer structure is used to obtain the local features of the disease area to help the CNN see better. Meanwhile, the paper designs different models according to the number of parameters and FLOPs (floating-point operations) to improve the model's scalability. The model variants of different sizes are designed to be lightweight to run on devices with different resource constraints. We achieved 98.78% identification accuracy on the self-built kiwifruit disease dataset, with up to 4.53% improvement in identification accuracy compared to the same level of Resnet, ViT, and ResMLP, and more than 10% reduction in the number of parameters and FLOPs. Experimental results on the PlantVillage dataset and the AI Challenger 2018 also show that ConvViT has good generalizability, indicating that the proposed model can solve kiwifruit disease identification problems in complex environments and be valuable a backbone network for other identification tasks with practical applications.

A centrifugal fan blade damage identification method based on the multi-level fusion of vibro-acoustic signals and CNN

A single sensor's blade damage identification is difficult because of the complex noise environment. At the same time, the multi-source signals include complete information of fault characteristics. Aiming to effectively fuse multi-sensor signals and improve identification accuracy, a centrifugal fan blade damage identification method based on the multi-level fusion of vibro-acoustic signals and a one-dimensional convolutional neural network (1D-CNN) is proposed. Firstly, acoustic and vibration signals are fused at the data level respectively by a data-adaptive synchronization weighted fusion algorithm. Secondly, the proposed 1D-CNN network extracts feature from the fused acoustic and vibration signals. Finally, the extracted features are fused by a fully connected layer. The experimental results show that the proposed method achieves 100% recognition accuracy at four speeds. By analyzing different signal to noise ratios (SNR), this method has higher diagnostic accuracy and better robustness compared with single sensor, single-level fusion, and other diagnostic methods.

ALNN‐based LOS/NLOS identification in 3D millimetre wave channel

Line-of-sight (LOS)/non-line-of-sight (NLOS) identification is crucial for millimetre-wave (mmW) since it is vulnerable to blocking in the NLOS environment. In this paper, an LOS/NLOS environment identification method is proposed based on angle information learning. Specifically, a neural network named ALNN (angle information learning neural network) is employed, which can learn the difference between the various angle paraments to identify the LOS and NLOS environments for mmW wireless communication systems. First, the 3D mmW channel model is applied to extract the desired angle information. Moreover, an mmW wireless transmission system is constructed to introduce actual measurement data to test the proposed method. Furthermore, various hyperparameters of the ALNN are adjusted to improve identification accuracy. Simulation results show that the proposed method has outstanding identification performance, and the identification accuracy is up to 99.41%.

Identification of Cultivated Land Quality Grade Using Fused Multi-Source Data and Multi-Temporal Crop Remote Sensing Information

To explore the fast, accurate, and efficient remote sensing identification method of cultivated land quality, this study took Shandong Province as the study area, and used measured data to carry out the soil quality evaluation based on conventional GIS. On this basis, MODIS sequence images were used as remote sensing data sources, and multi-source data such as topography, meteorology, and statistical yearbook were fused. Then, according to the Pressure-State-Response framework, we constructed three kinds of characteristic indicators through distinguishing crop rotation types and fusing remote sensing data. Finally, the soil quality grade was identified by the random forest method, and the accuracy analysis was carried out. The results showed that the NDVI peak values of double-season crops are in mid-April and mid-August, and one-season crops are in mid-August. Through evaluation, soil quality was divided into three categories, with six grades. Through principal component analysis, each soil status indicator contains two to three principal components, and each principal component contains five to eight temporal crop remote sensing information. After distinguishing crop rotation types and fusing remote sensing images, the identification accuracy of soil quality is significantly improved. The overall accuracy is 79.18%, 86.12%, and 93.65%, and the Kappa coefficient is 0.66, 0.77, and 0.90, respectively. This research developed an automatic identification method for cultivated land quality grade, and it proved that distinguishing crop rotation types and fusing multi-temporal crop remote sensing information are effective ways to improve identification accuracy.

Individual automatic detection and identification of big cats with the combination of different body parts.

The development of facial recognition technology has become an increasingly powerful tool in wild animal individual recognition. In this paper, we develop an automatic detection and recognition method with the combinations of body features of big cats based on the deep convolutional neural network (CNN). We collected dataset including 12244 images from 47 individual Amur tigers (Panthera tigris altaica) at the Siberian Tiger Park by mobile phones and digital camera and 1940 images and videos of 12 individual wild Amur leopard (Panthera pardus orientalis) by infrared cameras. First, the single shot multibox detector algorithm is used to perform the automatic detection process of feature regions in each image. For the different feature regions of the image, like face stripe or spots, CNNs and multi-layer perceptron models were applied to automatically identify tiger and leopard individuals, independently. Our results show that the identification accuracy of Amur tiger can reach up to 93.27% for face front, 93.33% for right body stripe, and 93.46% for left body stripe. Furthermore, the combination of right face, left body stripe, and right body stripe achieves the highest accuracy rate, up to 95.55%. Consequently, the combination of different body parts can improve the individual identification accuracy. However, it is not the higher the number of body parts, the higher the accuracy rate. The combination model with 3 body parts has the highest accuracy. The identification accuracy of Amur leopard can reach up to 86.90% for face front, 89.13% for left body spots, and 88.33% for right body spots. The accuracy of different body parts combination is lower than the independent part. For wild Amur leopard, the combination of face with body spot part is not helpful for the improvement of identification accuracy. The most effective identification part is still the independent left or right body spot part. It can be applied in long-term monitoring of big cats, including big data analysis for animal behavior, and be helpful for the individual identification of other wildlife species.

Attention Module Magnetic Flux Leakage Linked Deep Residual Network for Pipeline In-Line Inspection.

Pipeline operational safety is the foundation of the pipeline industry. Inspection and evaluation of defects is an important means of ensuring the safe operation of pipelines. In-line inspection of Magnetic Flux Leakage (MFL) can be used to identify and analyze potential defects. For pipeline MFL identification with inspecting in long distance, there exists the issues of low identification efficiency, misjudgment and leakage judgment. To solve these problems, a pipeline MFL inspection signal identification method based on improved deep residual convolutional neural network and attention module is proposed. A improved deep residual network based on the VGG16 convolution neural network is constructed to automatically learn the features from the MFL image signals and perform the identification of pipeline features and defects. The attention modules are introduced to reduce the influence of noises and compound features on the identification results in the process of in-line inspection. The actual pipeline in-line inspection experimental results show that the proposed method can accurately classify the MFL in-line inspection image signals and effectively reduce the influence of noises on the feature identification results with an average classification accuracy of 97.7%. This method can effectively improve identification accuracy and efficiency of the pipeline MFL in-line inspection.

Data-Driven Approach for Identifying Mistuning in As-Manufactured Blisks

Abstract Sector-to-sector geometry or material property variations in as-manufactured bladed disks, or blisks, can result in significantly greater vibration responses during operation compared to nominally cyclic symmetric designs. The dynamics of blisks are sensitive to these unavoidable deviations, known as mistuning, making the identification of mistuning in as-manufactured blisks necessary for accurately predicting their vibration. As in previous mistuning modeling and identification approaches, the mistuning of interest is small and is parameterized by using deviations in cantilever blade-alone frequencies. Such mistuning parameterization is popular because it can be applied through blade-to-blade stiffness deviations in computational reduced-order models used to predict blisk dynamics. Previous approaches to identify such mistuning parameters often require the identification of modal information or blade-isolation techniques such as blade detuning using masses or adding damping pads. However, modal information can be difficult to obtain accurately even in optimal bench conditions. In addition, in practice it can be difficult to isolate individual blades by restricting blade motion around the blisk or detuning individual blades through added masses due to geometric constraints. In this article, we present a method for mistuning identification using a data-driven approach based on a neural network. The network is first trained using surrogate computational data. Thus, the data-driven portion of the approach is executed using surrogate computational methods. With the trained network, mistuning in all sectors of blisks with the same nominal geometry can be identified by using a small number of forced responses and the forcing phase information from traveling-wave excitation. In this approach, no system or sector-level modal response information, restrictive blade isolation, or mass detuning are required. We additionally present a method for forcing frequency selection and response conditioning to improve identification accuracy. Validation of this approach is presented using a finite element blisk model containing stiffness mistuning within the blades to create computationally generated surrogate data. It is shown that mistuning can be predicted accurately using forced responses containing a significant amount of absolute and relative measurement noise, mimicking responses collected from experimental measurements. In addition, it is shown that mistuning can be predicted independently and accurately using different engine orders of excitation in regions of high modal density.