Deep Convolutional Neural Network Structure Research Articles

Motion coprime array-based DOA estimation considering phase disturbance of sensor array

To address the phase disturbance issue faced by sensor arrays in practical applications, a cascaded deep convolutional neural network structure is proposed to achieve direction-of-arrival (DOA) estimation for motion coprime arrays. Firstly, the synthesized covariance matrix obtained after motion is inputted into the first-level network for estimating the phase disturbance matrix. Then, we analyze the impact of phase perturbation on the synthesized covariance matrix and utilize the estimated disturbance phase to obtain an undisturbed synthesized covariance matrix. Finally, after phase compensation, the synthesized covariance matrix performs DOA estimation through the second-level network. Furthermore, to acquire three times the virtual unique lags of the coprime array, the synthesis condition about moving distance and unique lags is derived. The proposed method is shown to be effective and superior through the experiment results.

A CNN pruning approach using constrained binary particle swarm optimization with a reduced search space for image classification

Deep convolutional neural networks (CNNs) have exhibited exceptional performance in a range of computer vision tasks. However, these deep CNNs typically demand significant computational resources, which not only hinders their practical deployment but also contributes to a considerable carbon footprint. To tackle this issue, several filter pruning methods based on evolutionary algorithms have been proposed to provide significant memory and energy savings during CNN inference. However, due to the curse of high dimensionality in the structure of deep CNNs, the search space expands dramatically, presenting significant challenges for these methods. This paper proposes a novel algorithm called BPSO-FPruner for CNN filter pruning. BPSO-FPruner utilizes a constrained binary particle swarm optimization algorithm for filter pruning, incorporating a new initialization strategy based on filter weighting information and a reduced search space strategy. Extensive validation using VGG, ResNet, DenseNet, and MobileNetv2 architectures on the CIFAR-10, CIFAR-100, and Tiny ImageNet datasets demonstrates the effectiveness of BPSO-FPruner in reducing model computational costs and carbon footprint emissions while maintaining or improving performance.

Roadmap for providing and leveraging annotated data by cytologists in the PDAC domain as open data: support for AI-based pathology image analysis development and data utilization strategies.

Pancreatic cancer is one of the most lethal cancers worldwide, with a 5-year survival rate of less than 5%, the lowest of all cancer types. Pancreatic ductal adenocarcinoma (PDAC) is the most common and aggressive pancreatic cancer and has been classified as a health emergency in the past few decades. The histopathological diagnosis and prognosis evaluation of PDAC is time-consuming, laborious, and challenging in current clinical practice conditions. Pathological artificial intelligence (AI) research has been actively conducted lately. However, accessing medical data is challenging; the amount of open pathology data is small, and the absence of open-annotation data drawn by medical staff makes it difficult to conduct pathology AI research. Here, we provide easily accessible high-quality annotation data to address the abovementioned obstacles. Data evaluation is performed by supervised learning using a deep convolutional neural network structure to segment 11 annotated PDAC histopathological whole slide images (WSIs) drawn by medical staff directly from an open WSI dataset. We visualized the segmentation results of the histopathological images with a Dice score of 73% on the WSIs, including PDAC areas, thus identifying areas important for PDAC diagnosis and demonstrating high data quality. Additionally, pathologists assisted by AI can significantly increase their work efficiency. The pathological AI guidelines we propose are effective in developing histopathological AI for PDAC and are significant in the clinical field.

Impact of Deep Convolutional Neural Network Structure on Photovoltaic Array Extraction from High Spatial Resolution Remote Sensing Images

Accurate information on the location, shape, and size of photovoltaic (PV) arrays is essential for optimal power system planning and energy system development. In this study, we explore the potential of deep convolutional neural networks (DCNNs) for extracting PV arrays from high spatial resolution remote sensing (HSRRS) images. While previous research has mainly focused on the application of DCNNs, little attention has been paid to investigating the influence of different DCNN structures on the accuracy of PV array extraction. To address this gap, we compare the performance of seven popular DCNNs—AlexNet, VGG16, ResNet50, ResNeXt50, Xception, DenseNet121, and EfficientNetB6—based on a PV array dataset containing 2072 images of 1024 × 1024 size. We evaluate their intersection over union (IoU) values and highlight four DCNNs (EfficientNetB6, Xception, ResNeXt50, and VGG16) that consistently achieve IoU values above 94%. Furthermore, through analyzing the difference in the structure and features of these four DCNNs, we identify structural factors that contribute to the extraction of low-level spatial features (LFs) and high-level semantic features (HFs) of PV arrays. We find that the first feature extraction block without downsampling enhances the LFs’ extraction capability of the DCNNs, resulting in an increase in IoU values of approximately 0.25%. In addition, the use of separable convolution and attention mechanisms plays a crucial role in improving the HFs’ extraction, resulting in a 0.7% and 0.4% increase in IoU values, respectively. Overall, our study provides valuable insights into the impact of DCNN structures on the extraction of PV arrays from HSRRS images. These findings have significant implications for the selection of appropriate DCNNs and the design of robust DCNNs tailored for the accurate and efficient extraction of PV arrays.

A High-Performance Approach for Irregular License Plate Recognition in Unconstrained Scenarios

This paper proposes a novel framework for locating and recognizing irregular license plates in real-world complex scene images. In the proposed framework, an efficient deep convolutional neural network (CNN) structure specially designed for keypoint estimation is first employed to predict the corner points of license plates. Then, based on the predicted corner points, perspective transformation is performed to align the detected license plates. Finally, a lightweight deep CNN structure based on the YOLO detector is designed to predict license plate characters. The character recognition network can predict license plate characters without depending on license plate layouts (i.e., license plates of single-line or double-line text). Experiment results on CCPD and AOLP datasets demonstrate that the proposed method obtains better recognition accuracy compared with previous methods. The proposed model also achieves impressive inference speed and can be deployed in real-time applications.

Approximating functions with multi-features by deep convolutional neural networks

Deep convolutional neural networks (DCNNs) have achieved great empirical success in many fields such as natural language processing, computer vision, and pattern recognition. But there still lacks theoretical understanding of the flexibility and adaptivity of DCNNs in various learning tasks, and the power of DCNNs at feature extraction. We propose a generic DCNN structure consisting of two groups of convolutional layers associated with two downsampling operators, and a fully connected layer, which is determined only by three structural parameters. Our generic DCNNs are capable of extracting various features including not only polynomial features but also general smooth features. We also show that the curse of dimensionality can be circumvented by our DCNNs for target functions of the compositional form with (symmetric) polynomial features, spatially sparse smooth features, and interaction features. These demonstrate the expressive power of our DCNN structure, while the model selection can be relaxed comparing with other deep neural networks since there are only three hyperparameters controlling the architecture to tune.

Improving the Classification of Real-World SSVEP Data in Brain-Computer Interface Speller Systems Using Deep Convolutional Neural Networks

Purpose: Brain-Computer Interface (BCI) Speller systems help people with mobility impairments improve their cognitive and physical abilities. Steady-State Visual Evoked Potential (SSVEP) signals have been used to build high-speed BCI speller systems. SSVEP signals are a subtype of Visual Evoked Potential (VEP), a form of co-frequency, and the harmonics response elicited by a specific frequency stimulus. Noise and artifacts are critical issues for target detection in SSVEP-based BCI systems.
 Materials and Methods: Thus, it is essential to provide target detection techniques that operate well in the presence of noises. One solution for overcoming the noise impact is to employ approaches that automatically extract the appropriate features for target detection from the training data. Deep Convolutional Neural Network (DCNN) was utilized in this study to automatically extract features from SSVEP data in noisy conditions. Moreover, the BETA database, which contains SSVEP data from 70 individuals collected outside of the electromagnetic shielding room, was used. In this regard, a suitable DCNN structure for target stimulus frequency identification was first designed. The network was pre-trained with part of the data from the BETA database. Finally, at the single-subject level, this pre-trained network was retrained and evaluated.
 Results: The results showed that after retraining, the accuracy and Information Transfer Rate (ITR) increased (p-value < 0.01) for all participants.
 Conclusion:.The enhancement in accuracy and ITR are 25.72% and 43.10 bpm, respectively.

Hybrid U-Net-based deep learning model for volume segmentation of lung nodules in CT images.

Accurate segmentation of the lung nodule in computed tomography images is a critical component of a computer-assisted lung cancer detection/diagnosis system. However, lung nodule segmentation is a challenging task due to the heterogeneity of nodules. This study is to develop a hybrid deep learning (H-DL) model for the segmentation of lung nodules with a wide variety of sizes, shapes, margins, and opacities. A dataset collected from Lung Image Database Consortium image collection containing 847 cases with lung nodules manually annotated by at least two radiologists with nodule diameters greater than 7mm and less than 45mm was randomly split into 683 training/validation and 164 independent test cases. The 50% consensus consolidation of radiologists' annotation was used as the reference standard for each nodule. We designed a new H-DL model combining two deep convolutional neural networks (DCNNs) with different structures as encoders to increase the learning capabilities for the segmentation of complex lung nodules. Leveraging the basic symmetric U-shaped architecture of U-Net, we redesigned two new U-shaped deep learning (U-DL) models that were expanded to six levels of convolutional layers. One U-DL model used a shallow DCNN structure containing 16 convolutional layers adapted from the VGG-19 as the encoder, and the other used a deep DCNN structure containing 200 layers adapted from DenseNet-201 as the encoder, while the same decoder with only one convolutional layer at each level was used in both U-DL models, and we referred to them as the shallow and deep U-DL models. Finally, an ensemble layer was used to combine the two U-DL models into the H-DL model. We compared the effectiveness of the H-DL, the shallow U-DL and the deep U-DL models by deploying them separately to the test set. The accuracy of volume segmentation for each nodule was evaluated by the 3D Dice coefficient and Jaccard index (JI) relative to the reference standard. For comparison, we calculated the median and minimum of the 3D Dice and JI over the individual radiologists who segmented each nodule, referred to as M-Dice, min-Dice, M-JI, and min-JI. For the 164 test cases with 327 nodules, our H-DL model achieved an average 3D Dice coefficient of 0.750 ± 0.135 and an average JI of 0.617 ± 0.159. The radiologists' average M-Dice was 0.778 ± 0.102, and the average M-JI was 0.651 ± 0.127; both were significantly higher than those achieved by the H-DL model (p<0.05). The radiologists' average min-Dice (0.685 ± 0.139) and the average min-JI (0.537 ± 0.153) were significantly lower than those achieved by the H-DL model (p<0.05). The results indicated that the H-DL model approached the average performance of radiologists and was superior to the radiologist whose manual segmentation had the min-Dice and min-JI. Moreover, the average Dice and average JI achieved by the H-DL model were significantly higher than those achieved by the individual shallow U-DL model (Dice of 0.745 ± 0.139, JI of 0.611 ± 0.161; p<0.05) or the individual deep U-DL model alone (Dice of 0.739 ± 0.145, JI of 0.604 ± 0.163; p<0.05). Our newly developed H-DL model outperformed the individual shallow or deep U-DL models. The H-DL method combining multilevel features learned by both the shallow and deep DCNNs could achieve segmentation accuracy comparable to radiologists' segmentation for nodules with wide ranges of image characteristics.

Wafer map failure pattern recognition based on deep convolutional neural network

The objective of this paper is to propose a systematic failure pattern recognition for wafer map based on neural networks. A deep convolutional neural network (DCNN) model which includes the convolutional layer, batch normalization layer, Relu layer, maximum pooling layer, full connection layer, Softmax layer and classification layer is established for the problem of failure pattern recognition of wafer map. This model is an end-to-end 19-layer network, which can actively learn and automatically extract effective classification features. After grayscale and median filtering, the wafer map can be imported into the network for automatic failure classification without special feature extraction. On this basis, we build a dual-source DCNN structure by combining decision-level information entropy fusion. The verification results of the model in the actual wafer map database WM-811K show that the model exhibits good performance in identifying nine kinds of common failure patterns, and has advantages in identifying non-pattern wafer patterns without failure patterns. The dual-source DCNN structure has a better classification effect than the single-source DCNN structure, and the overall recognition accuracy of the dual-source DCNN structure reaches 98.34%.

Modified layer deep convolution neural network for text-independent speaker recognition

peaker recognition is the task of identifying the spokesman automatically using speaker-specific features. It has been a popular and most involved topic in the field of speech technology. This field opens a wide opportunity for research and finds its application in the areas such as forensics, authentication, security, etc. In this work, a modified deep-convolutional neural network structure has been proposed for speaker identification that has improved convolution, activation, and pooling layers along with Adam’s optimiser. The proposed architecture yielded the increase of prediction accuracy and reduction of Loss function when compared to the generic Convolutional Neural Network scheme. The execution of the proposed architecture is validated by various datasets and the outcomes show that the modified CNN performs better than the other state-of-the-art models regarding both accuracy (avg 99%) and loss function (avg 1%). From the analysis, it is found that the Modified-CNN suits the best for real-time speaker identification applications as the efficacy of the model does not degrade due to the effects of noise and interferences that are caused in the recording environment. Relevance of the work: Speaker Recognition is an area of interest in which ML and DL schemes, when combined, have the potential to make history in the areas of Automation and Authentication. Using a modified CNN can enhance the process by ignoring many issues such as false positives, background noise, and so on. This process can be expanded to create a Raga Identification and Therapy mechanism that can be used to treat diseases.

Diagnose Parkinson’s disease and cleft lip and palate using deep convolutional neural networks evolved by IP-based chimp optimization algorithm

Speech signals often include paralinguistic features such as pathologies that impair a speaker’s capability to communicate. Those cognitive symptoms have various causes depending on the disease. For example, morphological diseases like cleft lip and palate create hypernasality, while neurodegenerative conditions like Parkinson’s disease cause hypokinetic dysarthria. Automatic assessment of abnormal speech supports early diagnosis or disease severity evaluation. Conventional methods rely on manually assessing single aspects like shimmer, jitter, or formant frequencies, which may not fully reflect the disease’s manifestations. In this paper, we use deep convolutional neural networks (DCNNs) to recognize disordered speech. Despite DCNNs’ many approved benefits, selecting the best structure is challenging. In order to overcome this issue, this research looks into using the chimp optimization algorithm (ChOA) to automatically select the optimal DCNN structure. In order to achieve the goal, three ChOA-based advancements are proposed. First, an internet protocol address-based (IPA-based) encoding method for DCNN layers employing chimp vectors is created. Then an Enfeebled layer with specified chimp vector dimensions is presented for variable-length DCNNs. Eventually, large datasets are partitioned into smaller ones and evaluated at random to recognize abnormal speech signals from patients with Parkinson’s disease and cleft lip and palate. In addition to receiver operating characteristic (ROC) and precision-recall curves, five well-known metrics were used: sensitivity, specificity, accuracy, precision, F1-Score. The proposed model accurately diagnoses disordered and normal speech signals, with an accuracy of up to 96.37%, which is 1.62 more accurate than the second-best approach, i.e., VLNSGA-II.

An Approach for Optimization of Features using Gorilla Troop Optimizer for Classification of Melanoma

The diagnosis and categorization of skin cancer, as well as the difference in skin textures and injuries, is a tough undertaking. Manually detecting skin lesions from dermoscopy images seems to be a difficult and cumbersome challenge. Recent advancements in the internet of things (IoT) and artificial intelligence for clinical applications have shown significant increase in precision and processing time. A lot of attention is given to deep learning models because they are effective at identifying cancer cells. The diagnosis and accuracy levels can be greatly increased by categorizing benign and malignant dermoscopy images. This work suggests an automated classification system based on a deep convolutional neural network (DCNN) in order to precisely perform multi-classification. The DCNN's structure was thoughtfully created by arranging a number of layers that are in charge of uniquely extracting different features from skin lesions. In this paper, we proposed a deep learning approach to tackle the three main tasks-deep extraction of features (task1) using transfer learning, selection of features (task2)-using metaheuristic algorithms such as Particle Swarm Optimization (PSO), Ant Colony Optimization (ACO), and Gorilla Troop Optimization (GTO) as a feature selector, the extensive feature set is optimized, and the amount of features is reduced to within the range, and a two-level classification (task3) was proposed that are emerging in the field of skin lesion image processing. On the HAM10000 dataset, the proposed deep learning frameworks were assessed. The accuracy achieved on the dataset is 93.58 percent. The proposed method outperforms state-of-the-art (SOTA) techniques in terms of accuracy. The suggested technique is however highly scalable.

PO-1179 Radiation induced pneumonitis during COVID-19: artificial intelligence for differential diagnosis.

Purpose or Objective: In 2019, the Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) was identified in Wuhan, China and in March 2020 the World Health Organization (WHO) declared the global public health emergency describing the situation as a pandemic. The most serious clinical entity of the respiratory syndrome associated with SARS-CoV-2 is a severe interstitial pneumonia. Radiation pneumonitis (RP) is a typical toxicity related to chemoradiation for locally advanced lung cancer patients. RP and SARS-CoV-2 interstitial pneumonia show overlapping clinical features and differential diagnosis maybe be challenging. The aim of this study is to test the performance of a deep learning algorithm in discriminating radiation pneumonitis (RP) from COVID-19 pneumonia. Materials and Methods: Seventy patients were analysed, thirty-four affected by COVID-19 pneumonia and thirty-six by radiation therapy-related pneumonitis (RP group). The CT images were quantitatively analyzed by InferReadTM CT Lung (COVID-19) (Infervision, Europe GmbH, Wiesbaden, Germany), an Artificial Intelligence solution specifically developed for diagnosis and management support of COVID-19 pneumonia, based on an AI algorithm built on a novel deep convolutional neural network structure. Based on a preliminary analysis of the deep-learning algorithm, the cut-off value of the estimated risk probability of COVID-19 was set at levels higher than 30% (“COVID19 High Risk”), as the percentage of COVID-19 confirmed patients above this cut-off value was higher than 95%. Values of estimated risk probability below 30% were classified as “COVID19 Low Risk. Results: Most patients presenting RP were classified by the algorithm as “COVID19 Low Risk” (66.7%). All RP classified as “COVID19 High Risk” were ≥G3 (CTC AE vers. 4.0). The algorithm showed good accuracy in the detection of RP against COVID-19 pneumonia (sensitivity = 97.0%, specificity = 2%, AUC = 0.72). This accuracy increased when an estimated COVID-19 risk probability cut-off of 30% was applied (sensitivity 76%, specificity 63%, AUC = 0.84). The total lung volume involvement was higher in COVID 19 patients compared with RP group (mean= 105.54 cc, IQ range= 44.68-257.07 vs mean=29.14 cc, IQ range= 5.59-69.20, p <0.001). In patients pretreated with radiation therapy and actually presenting diffuse pneumonitis classified by AI as “COVID19 High Risk” a combination of dosimetric factors may help to identify RP (PPV increased from 60% to 99.8%). Conclusion: Deep-learning algorithm can help to discriminate RP from COVID-19 pneumonia, classifying most RP as “Lowrisk COVID19” (below the cut off value of COVID-19 risk probability of 30%). In patients classified as high risk , treated with radiation therapy also dosimetric factors should be taken into account.

Radiation-Induced Pneumonitis in the Era of the COVID-19 Pandemic: Artificial Intelligence for Differential Diagnosis.

Simple SummaryRadiation-induced pneumonitis and severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) interstitial pneumonia show overlapping clinical features. As we are facing the COVID-19 pandemic, the discrimination between these two entities is of paramount importance. In fact, lung cancer patients are at higher risk of complications from SARS-CoV-2. In this study, we aimed to investigate if a deep learning algorithm was able to discriminate between COVID-19 and radiation therapy-related pneumonitis (RP). The algorithm showed high sensitivity but low specificity in the detection of RP against COVID-19 pneumonia (sensitivity = 97.0%, specificity = 2%, area under the curve (AUC = 0.72). The specificity increased when an estimated COVID-19 risk probability cut-off of 30% was applied (sensitivity 76%, specificity 63%, AUC = 0.84).(1) Aim: To test the performance of a deep learning algorithm in discriminating radiation therapy-related pneumonitis (RP) from COVID-19 pneumonia. (2) Methods: In this retrospective study, we enrolled three groups of subjects: pneumonia-free (control group), COVID-19 pneumonia and RP patients. CT images were analyzed by mean of an artificial intelligence (AI) algorithm based on a novel deep convolutional neural network structure. The cut-off value of risk probability of COVID-19 was 30%; values higher than 30% were classified as COVID-19 High Risk, and values below 30% as COVID-19 Low Risk. The statistical analysis included the Mann–Whitney U test (significance threshold at p < 0.05) and receiver operating characteristic (ROC) curve, with fitting performed using the maximum likelihood fit of a binormal model. (3) Results: Most patients presenting RP (66.7%) were classified by the algorithm as COVID-19 Low Risk. The algorithm showed high sensitivity but low specificity in the detection of RP against COVID-19 pneumonia (sensitivity = 97.0%, specificity = 2%, area under the curve (AUC = 0.72). The specificity increased when an estimated COVID-19 risk probability cut-off of 30% was applied (sensitivity 76%, specificity 63%, AUC = 0.84). (4) Conclusions: The deep learning algorithm was able to discriminate RP from COVID-19 pneumonia, classifying most RP cases as COVID-19 Low Risk.

Plant taxonomy-guided path-based tree classifier for large-scale plant species identification

A deep learning framework is proposed to recognize large-scale plant species by integrating attention-based deep feature extraction network and plant taxonomy-guided path-based tree classifier. First, a plant taxonomy is constructed for organizing large-scale fine-grained plant species hierarchically in a coarse-to-fine fashion. Second, a deep learning framework is proposed, where attention mechanism is used to remove useless feature components and a plant taxonomy-guided path-based two-layer tree classifier is used to replace the flat softmax classifier in traditional deep convolutional neural network structure. Furthermore, a specific path-based loss function and back-propagation method are proposed to optimize the weight parameters in both deep network and tree classifier. Experimental results on the Orchid2608 plant dataset can also prove that proposed deep attention network with path-based tree classifier can achieve improvements on large-scale plant species identification task.

Deep Learning Algorithm Trained with COVID-19 Pneumonia Also Identifies Immune Checkpoint Inhibitor Therapy-Related Pneumonitis.

Simple SummaryThe use of immune checkpoint inhibitors (ICIs) to treat oncologic diseases is progressively increasing. Computed tomography (CT) features of ICI therapy-related pneumonitis may overlap with other diseases, including coronavirus disease 2019 (COVID-19). Thus, oncologic patients undergoing ICI therapy and developing pneumonitis are at risk of being misdiagnosed. Exploring the strengths and weaknesses of artificial intelligence in distinguishing between ICI therapy-related pneumonitis and COVID-19 is of great importance for oncologic patients and for clinicians in order to increase awareness on this topic and stimulate novel strategies aimed to promptly and correctly classify and treat this category of vulnerable patients.Background: Coronavirus disease 2019 (COVID-19) pneumonia and immune checkpoint inhibitor (ICI) therapy-related pneumonitis share common features. The aim of this study was to determine on chest computed tomography (CT) images whether a deep convolutional neural network algorithm is able to solve the challenge of differential diagnosis between COVID-19 pneumonia and ICI therapy-related pneumonitis. Methods: We enrolled three groups: a pneumonia-free group (n = 30), a COVID-19 group (n = 34), and a group of patients with ICI therapy-related pneumonitis (n = 21). Computed tomography images were analyzed with an artificial intelligence (AI) algorithm based on a deep convolutional neural network structure. Statistical analysis included the Mann–Whitney U test (significance threshold at p < 0.05) and the receiver operating characteristic curve (ROC curve). Results: The algorithm showed low specificity in distinguishing COVID-19 from ICI therapy-related pneumonitis (sensitivity 97.1%, specificity 14.3%, area under the curve (AUC) = 0.62). ICI therapy-related pneumonitis was identified by the AI when compared to pneumonia-free controls (sensitivity = 85.7%, specificity 100%, AUC = 0.97). Conclusions: The deep learning algorithm is not able to distinguish between COVID-19 pneumonia and ICI therapy-related pneumonitis. Awareness must be increased among clinicians about imaging similarities between COVID-19 and ICI therapy-related pneumonitis. ICI therapy-related pneumonitis can be applied as a challenge population for cross-validation to test the robustness of AI models used to analyze interstitial pneumonias of variable etiology.

Deep Scattering Network With Fractional Wavelet Transform

Deep convolutional neural networks (DCNNs) have recently emerged as a powerful tool to deliver breakthrough performances in various image analysis and processing applications. However, DCNNs lack a strong theoretical foundation and require massive amounts of training data. More recently, the deep scattering network (DSN), a variant of DCNNs, has been proposed to address these issues. DSNs inherit the hierarchical structure of DCNNs, but replace data-driven linear filters with predefined fixed multi-scale wavelet filters, which facilitate an in-depth understanding of DCNNs and also offer the state-of-the-art performance in image classification. Unfortunately, DSNs suffer from a major drawback: they are suitable for stationary image textures but not non-stationary image textures, since 2D wavelets are intrinsically linear translation-invariant filters in the Fourier transform domain. The objective of this paper is to overcome this drawback using the fractional wavelet transform (FRWT) which can be viewed as a bank of linear translation-variant multi-scale filters and thus may be well suited for non-stationary texture analysis. We first propose the fractional wavelet scattering transform (FRWST) based upon the FRWT. Then, we present a generalized structure for the DSN by cascading fractional wavelet convolutions and modulus operators. Basic properties of this generalized DSN are derived, followed by a fast implementation of the generalized DSN as well as their practical applications. The theoretical derivations are validated via computer simulations.

CoronaNet: A Novel Deep Learning Model for COVID-19 Detection in CT Scans

Coronavirus disease (COVID-19) is currently the cause of a global pandemic that is affecting millions of people around the world. Inadequate testing resources have resulted in several people going undiagnosed and consequently untreated; however, using computerized tomography (CT) scans for diagnosis is an alternative to bypass this limitation. Unfortunately, CT scan analysis is time-consuming and labor intensive and rendering is generally infeasible in most diagnosis situations. In order to alleviate this problem, previous studies have utilized multiple deep learning techniques to analyze biomedical images such as CT scans. Specifically, convolutional neural networks (CNNs) have been shown to provide medical diagnosis with a high degree of accuracy. A common issue in the training of CNNs for biomedical applications is the requirement of large datasets. In this paper, we propose the use of affine transformations to artificially magnify the size of our dataset. Additionally, we propose the use of the Laplace filter to increase feature detection in CT scan analysis. We then feed the preprocessed images to a novel deep CNN structure: CoronaNet. We find that the use of the Laplace filter significantly increases the performance of CoronaNet across all metrics. Additionally, we find that affine transformations successfully magnify the dataset without resulting in high degrees of overfitting. Specifically, we achieved an accuracy of 92% and an F1 of 0.8735. Our novel research describes the potential of the Laplace filter to significantly increase deep CNN performance in biomedical applications such as COVID-19 diagnosis.

Deep convolutional neural network structure design for remote sensing image scene classification based on transfer learning

To obtain an ideal scene classification effect when applying a deep convolutional neural network (DCNN) to remote sensing images, a DCNN named ConNet-3F based on transfer learning was constructed. Firstly, the initial 13 layers of VGG16 were used as the main architecture, and then 6 layers were added to the end of it to extract deeper features. The pruned model of VGG16 acted as the feature extractor to extract features from remote sensing images. Subsequently, the parameters trained on the ImageNet were set as initialization. Finally, parameters of all layers were trained on the remote sensing data to obtain the favorable model. The results on the Moving and Stationary Target Acquisition and Recognition (MSTAR) dataset demonstrate that this method achieves higher classification accuracy compared with other mentioned methods. Especially, even if the ratio of training data was reduced on the NWPU-RESISC45 dataset, classification accuracies of the proposed method always stay above 90%.

Major Depressive Disorder Classification Based on Different Convolutional Neural Network Models: Deep Learning Approach.

The human brain is characterized by complex structural, functional connections that integrate unique cognitive characteristics. There is a fundamental hurdle for the evaluation of both structural and functional connections of the brain and the effects in the diagnosis and treatment of neurodegenerative diseases. Currently, there is no clinically specific diagnostic biomarker capable of confirming the diagnosis of major depressive disorder (MDD). Therefore, exploring translational biomarkers of mood disorders based on deep learning (DL) has valuable potential with its recently underlined promising outcomes. In this article, an electroencephalography (EEG)-based diagnosis model for MDD is built through advanced computational neuroscience methodology coupled with a deep convolutional neural network (CNN) approach. EEG recordings are analyzed by modeling 3 different deep CNN structure, namely, ResNet-50, MobileNet, Inception-v3, in order to dichotomize MDD patients and healthy controls. EEG data are collected for 4 main frequency bands (Δ, θ, α, and β, accompanying spatial resolution with location information by collecting data from 19 electrodes. Following the pre-processing step, different DL architectures were employed to underline discrimination performance by comparing classification accuracies. The classification performance of models based on location data, MobileNet architecture generated 89.33% and 92.66% classification accuracy. As to the frequency bands, delta frequency band outperformed compared to other bands with 90.22% predictive accuracy and area under curve (AUC) value of 0.9 for ResNet-50 architecture. The main contribution of the study is the delineation of distinctive spatial and temporal features using various DL architectures to dichotomize 46 MDD subjects from 46 healthy subjects. Exploring translational biomarkers of mood disorders based on DL perspective is the main focus of this study and, though it is challenging, with its promising potential to improve our understanding of the psychiatric disorders, computational methods are highly worthy for the diagnosis process and valuable in terms of both speed and accuracy compared with classical approaches.