Types Of Convolutional Neural Networks Research Articles

Semantic segmentation-based detection algorithm for challenging cryo-electron microscopy RNP samples.

In this study, we present a novel and robust methodology for the automatic detection of influenza A virus ribonucleoproteins (RNPs) in single-particle cryo-electron microscopy (cryo-EM) images. Utilizing a U-net architecture-a type of convolutional neural network renowned for its efficiency in biomedical image segmentation-our approach is based on a pretraining phase with a dataset annotated through visual inspection. This dataset facilitates the precise identification of filamentous RNPs, including the localization of the filaments and their terminal coordinates. A key feature of our method is the application of semantic segmentation techniques, enabling the automated categorization of micrograph pixels into distinct classifications of particle and background. This deep learning strategy allows to robustly detect these intricate particles, a crucial step in achieving high-resolution reconstructions in cryo-EM studies. To encourage collaborative advancements in the field, we have made our routines, the pretrained U-net model, and the training dataset publicly accessible. The reproducibility and accessibility of these resources aim to facilitate further research and validation in the realm of cryo-EM image analysis.

Motor Fault Diagnosis Based on Convolutional Block Attention Module-Xception Lightweight Neural Network.

Electric motors play a crucial role in self-driving vehicles. Therefore, fault diagnosis in motors is important for ensuring the safety and reliability of vehicles. In order to improve fault detection performance, this paper proposes a motor fault diagnosis method based on vibration signals. Firstly, the vibration signals of each operating state of the motor at different frequencies are measured with vibration sensors. Secondly, the characteristic of Gram image coding is used to realize the coding of time domain information, and the one-dimensional vibration signals are transformed into grayscale diagrams to highlight their features. Finally, the lightweight neural network Xception is chosen as the main tool, and the attention mechanism Convolutional Block Attention Module (CBAM) is introduced into the model to enforce the importance of the characteristic information of the motor faults and realize their accurate identification. Xception is a type of convolutional neural network; its lightweight design maintains excellent performance while significantly reducing the model's order of magnitude. Without affecting the computational complexity and accuracy of the network, the CBAM attention mechanism is added, and Gram's corner field is combined with the improved lightweight neural network. The experimental results show that this model achieves a better recognition effect and faster iteration speed compared with the traditional Convolutional Neural Network (CNN), ResNet, and Xception networks.

Image cropping for malaria parasite detection on heterogeneous data

Malaria is a deadly disease of significant concern for the international community. It is an infectious disease caused by a Plasmodium spp. parasite and transmitted by the bite of an infected female Anopheles mosquito. The parasite multiplies in the liver and then destroys the person's red blood cells until it reaches the severe stage, leading to death. The most used tools for diagnosing this disease are the microscope and the rapid diagnostic test (RDT), which have limitations preventing control of the disease. Computer vision technologies present alternatives by providing the means for early detection of this disease before it reaches the severe stage, facilitating treatment and saving patients. In this article, we suggest deep learning methods for earlier and more accurate detection of malaria parasites with high generalization capabilities using microscopic images of blood smears from many heterogeneous patients. These techniques are based on an image preprocessing method that mitigates some of the challenges associated with the variety of red cell characteristics due to patient diversity and other artifacts present in the data.For the study, we collected 65,970 microscopic images from 876 different patients to form a dataset of 33,007 images with a variety that enables us to create models with a high level of generalization. Three types of convolutional neural networks were used, namely Convolutional Neural Network (CNN), DenseNet, and LeNet-5, and the highest classification accuracy on the test data was 97.50% found with the DenseNet model.

Classification of brain tumor types through MRIs using parallel CNNs and firefly optimization

Image segmentation is a critical and challenging endeavor in the field of medicine. A magnetic resonance imaging (MRI) scan is a helpful method for locating any abnormal brain tissue these days. It is a difficult undertaking for radiologists to diagnose and classify the tumor from several pictures. This work develops an intelligent method for accurately identifying brain tumors. This research investigates the identification of brain tumor types from MRI data using convolutional neural networks and optimization strategies. Two novel approaches are presented: the first is a novel segmentation technique based on firefly optimization (FFO) that assesses segmentation quality based on many parameters, and the other is a combination of two types of convolutional neural networks to categorize tumor traits and identify the kind of tumor. These upgrades are intended to raise the general efficacy of the MRI scan technique and increase identification accuracy. Using MRI scans from BBRATS2018, the testing is carried out, and the suggested approach has shown improved performance with an average accuracy of 98.6%.

Efficient Method of Road Outlier Recognition Using Deep Learning Coupled with Data Augmentation Approach

In recent years, Closed-circuit Television (CCTV) cameras have been playing a vital role in the surveillance of both public and private areas. The primary objective of surveillance is to monitor human behavior and road conditions. In the real-world situation, detecting, and recognizing abnormal activities poses significant challenges due to the densely crowded environment and the complex nature of transportation systems. These factors make it difficult to automatically identify various anomalies that occur while traveling, leading to emergencies, and endangering human life and property. In response to these challenges, contemporary Road Outlier Recognition Surveillance Systems (RORSS) have been deployed to monitor and mitigate abnormalities on streets, highways, and roads. This research specifically investigated conventional as well as human-centered road anomalies encompassing snatching, accidents, fighting, and car fires. The proposed Real-time Road Outlier Recognition methodology is formulated as a classification task, involving the analysis and processing of real-time CCTV videos. The proposed study investigated different types of Convolutional Neural Network (CNN) pre-trained models with the Data Augmentation (DA) approach to address the frame variance problem in real-time videos. Furthermore, we introduced a benchmark real-world Road Outlier Dataset (ROD) containing roads, streets, and highways videos and images that presented different road anomalies performed by vehicles and humans. Experimentation using different pre-trained CNN models i.e., VGG19, InceptionV3, ResNet50, and DenseNet201 with a data augmentation approach has been performed on the ROD dataset. The experimental results demonstrated that the InceptionV3 model performed best with the augmentation approach as compared to other deep learning models that achieved the best accuracy of 98.80% in recognizing road anomalies

A Method for Underwater Acoustic Target Recognition Based on the Delay-Doppler Joint Feature

With the aim of solving the problem of identifying complex underwater acoustic targets using a single signal feature in the Time–Frequency(TF) feature, this paper designs a method that recognizes the underwater targets based on the Delay-Doppler joint feature. First, this method uses symplectic finite Fourier transform (SFFT) to extract the Delay-Doppler features of underwater acoustic signals, analyzes the Time–Frequency features at the same time, and combines the Delay-Doppler (DD) feature and Time–Frequency feature to form a joint feature (TF-DD). This paper uses three types of convolutional neural networks to verify that TF-DD can effectively improve the accuracy of target recognition. Secondly, this paper designs an object recognition model (TF-DD-CNN) based on joint features as input, which simplifies the neural network’s overall structure and improves the model’s training efficiency. This research employs ship-radiated noise to validate the efficacy of TF-DD-CNN for target identification. The results demonstrate that the combined characteristic and the TF-DD-CNN model introduced in this study can proficiently detect ships, and the model notably enhances the precision of detection.

Object Detection of BISINDO Sign Language Letters Using Residual Network

Indonesian Sign Language or BISINDO is an alternative language used by people who suffer from disabilities, especially those who have hearing impairments. This language grew and developed from the deaf community, so its use is based on the visual aspect. This research aims to apply Residual Networks to detect objects in the context of Bisindo Letter Sign Language, with the hope of increasing accuracy and efficiency in letter recognition. Object detection goes through 2 stages, namely feature extraction and model training. ResNet is a type of Convolutional Neural Network (CNN) architecture that utilizes models that have been previously trained, so it can save the time required in the model development process. In this research, Residual Network (ResNet) was used for feature extraction to recognize important aspects in the Bisindo letter sign image, such as hand position, finger shape characteristics, and direction of movement. The research results show that the new dataset used as training data and test data has a fairly good ability to detect with a division of 70% train set, 20% valid set and 10% test set with size 640x640 with 300 epochs for the training model.

Implementation on FingerScripter: Converting Fingerspelling into Text and Speech

Abstract: Communication might seem a simple act for normal people, but for the ones who are suffering from physical or psychological disability find it very difficult for communicating with others. This communication barrier needs to be removed for which there are various techniques that are being used, out of which some are verbal, non-verbal, visual and written. For mute and deaf people these communication barrier is a major problem which can adversely affect their lives adding to their already challenging lives. For this sign languages are the best possible option, as they bridge the gap between communication of these deaf and mute people with normal people. We have created a machine learning model in order to solve this issue which will identify these signs and interpret them to detect which alphabet is signified. The model will be detecting the American Sign Language (ASL) which is one of the most widely used sign language and a type of fingerspelling. The model will form words out of the identified alphabets which will also be then converted into speech. The key algorithm which will be used is a special type of Convolutional Neural Network (CNN) for image recognition. All the documentation regarding the model along with its working is provided in this paper. With high precision and accuracy the model aims to ease the communication of deaf and mute people.

Wide-TSNet: A Novel Hybrid Approach for Bitcoin Price Movement Classification

In this paper, we introduce Wide-TSNet, a novel hybrid approach for predicting Bitcoin prices using time-series data transformed into images. The method involves converting time-series data into Markov transition fields (MTFs), enhancing them using histogram equalization, and classifying them using Wide ResNets, a type of convolutional neural network (CNN). We propose a tripartite classification system to accurately represent Bitcoin price trends. In addition, we demonstrate the effectiveness of Wide-TSNet through various experiments, in which it achieves an Accuracy of approximately 94% and an F1 score of 90%. It is also shown that lightweight CNN models, such as SqueezeNet and EfficientNet, can be as effective as complex models under certain conditions. Furthermore, we investigate the efficacy of other image transformation methods, such as Gramian angular fields, in capturing the trends and volatility of Bitcoin prices and revealing patterns that are not visible in the raw data. Moreover, we assess the effect of image resolution on model performance, emphasizing the importance of this factor in image-based time-series classification. Our findings explore the intersection between finance, image processing, and deep learning, providing a robust methodology for financial time-series classification.

Convolutional Neural Network Approaches in Median Nerve Morphological Assessment from Ultrasound Images.

Ultrasound imaging has been used to investigate compression of the median nerve in carpal tunnel syndrome patients. Ultrasound imaging and the extraction of median nerve parameters from ultrasound images are crucial and are usually performed manually by experts. The manual annotation of ultrasound images relies on experience, and intra- and interrater reliability may vary among studies. In this study, two types of convolutional neural networks (CNNs), U-Net and SegNet, were used to extract the median nerve morphology. To the best of our knowledge, the application of these methods to ultrasound imaging of the median nerve has not yet been investigated. Spearman's correlation and Bland-Altman analyses were performed to investigate the correlation and agreement between manual annotation and CNN estimation, namely, the cross-sectional area, circumference, and diameter of the median nerve. The results showed that the intersection over union (IoU) of U-Net (0.717) was greater than that of SegNet (0.625). A few images in SegNet had an IoU below 0.6, decreasing the average IoU. In both models, the IoU decreased when the median nerve was elongated longitudinally with a blurred outline. The Bland-Altman analysis revealed that, in general, both the U-Net- and SegNet-estimated measurements showed 95% limits of agreement with manual annotation. These results show that these CNN models are promising tools for median nerve ultrasound imaging analysis.

Redundant node identification using Siamese neural network to enhance the availability in visual sensor network

SummaryMany monitoring applications related to surveillance, tracking and multipurpose visual monitoring have taken into consideration the use of wireless visual sensor networks. When sensors are deployed over a monitored field that could potentially damage the monitoring capability and availability in the visual sensor network. In order to overcome these issues, faulty nodes are identified and replaced by redundant node based on Siamese network in this research. Initially, camera nodes are randomly deployed in the visual sensor network, and the data are received from the network through gateway. To identify the redundant nodes, initially, the frames are divided into equivalent time slot, and then, Siamese network is utilized to identify the redundant nodes in a network. Siamese neural network is type of convolutional neural network that is utilized to recognize the similar images in the network. After that, faulty nodes are identified based on some parameters such as entropy, energy, transmission delay and network coverage. If the average energy, entropy, transmission delay and network coverage are below the threshold value, then the node is identified as faulty node. Finally, replace the faulty node with redundant node to enhance the availability in the visual sensor network for critical monitoring applications. The simulation analysis shows that the developed approach takes 772 s to identify redundant node and take 27 s to identify fake nodes and the developed method is executed at 1308 s. Thus, this prediction model helps to improve the coverage quality of target‐based monitoring in order to achieve availability.

Time-Constrained Ensemble Sensing With Heterogeneous IoT Devices in Intelligent Transportation Systems

Recently we have witnessed the rise of Artificial Intelligence of Things (AIoT) and the shift of sensing paradigm from cloud-centric to the edge-centric, which effectively improves the sensing capability of intelligence transportation systems. To improve the real-time sensing performance, in this work we propose an ensemble sensing based scheme to solve the time-constraint synchronized inference problem and achieve robust inference with heterogeneous IoT devices in intelligence transportation systems. We design and implement Ensen, which incorporates various novel techniques such as customized DNN model design, KD-based model training, and dynamic deep ensemble management, etc., to achieve improved accuracy and maximize the computational resource usage of the whole sensing group. Extensive evaluations on different types of common IoT devices have shown that Ensen achieves a robust performance and can be easily extended to different types of convolutional neural networks.

Feature recognition in multiple CNNs using sEMG images from a prototype comfort test

ObjectiveDeep learning-based CNN networks have recently been investigated to solve the problem of body posture recognition based on surface electromyographic signals (sEMG). Influenced by these studies, to develop a combined approach of sEMG and CNNs in the study of human-product interactions and the impact of body comfort, and to compare the advantages and disadvantages of various CNNs networks. MethodsIn this study, sEMG measurements were carried out by building a prototype usability experiment, and the data were divided into four categories, with two types of datasets: training and testing. Four CNNs, LeNet-5, VGGNet-11, InceptionNet V4, and DenseNet, were used for the recognition of sEMG images. ResultsDenseNet is another type of convolutional neural network with deep layers, which has a unique advantage over other algorithms. unique advantages over other algorithms. DenseNet has fewer layers and better accuracy than InceptionNet V4, but not only does it bypass enhanced feature reuse, but its network is easier to train and has some regularization effects, while also mitigating the problems of gradient disappearance and model degradation. ConclusionThese findings could lead to a more appropriate CNN model and a useful tool for developing comfort judgments of surface EMG signals, furthering the development of products that come into contact with the human body without the need for routine retraining.

Florets for Chiplets: Data Flow-aware High-Performance and Energy-efficient Network-on-Interposer for CNN Inference Tasks

Recent advances in 2.5D chiplet platforms provide a new avenue for compact scale-out implementations of emerging compute- and data-intensive applications including machine learning. Network-on-Interposer (NoI) enables integration of multiple chiplets on a 2.5D system. While these manycore platforms can deliver high computational throughput and energy efficiency by running multiple specialized tasks concurrently, conventional NoI architectures have a limited computational throughput due to their inherent multi-hop topologies. In this paper, we propose Floret, a novel NoI architecture based on space-filling curves (SFCs). The Floret architecture leverages suitable task mapping, exploits the data flow pattern, and optimizes the inter-chiplet data exchange to extract high performance for multiple types of convolutional neural network (CNN) inference tasks running concurrently. We demonstrate that the Floret architecture reduces the latency and energy up to 58% and 64%, respectively, compared to state-of-the-art NoI architectures while executing datacenter-scale workloads involving multiple CNN tasks simultaneously. Floret achieves high performance and significant energy savings with much lower fabrication cost by exploiting the data-flow awareness of the CNN inference tasks.

Deep Learning-Based Identification of Common Complication Features of Surgical Incisions

In recent years, due to the development of accelerated recovery after surgery and day surgery in the field of surgery, the average length-of-stay of patients has been shortened and patients stay at home for post-surgical recovery and healing of the surgical incisions. In order to identify, in a timely manner, the problems that may appear at the incision site and help patients prevent or reduce the anxiety they may experience after discharge, we used deep learning method in this study to classify the features of common complications of surgical incisions, hoping to realize patient-directed early identification of complications common to surgical incisions. A total of 1 224 postoperative photographs of patients' surgical incisions were taken and collected at a tertiary-care hospital between June 2021 and March 2022. The photographs were collated and categorized according to different features of complications of the surgical incisions. Then, the photographs were divided into training, validation, and test sets at the ratio of 8∶1∶1 and 4 types of convolutional neural networks were applied in the training and testing of the models. Through the training of multiple convolutional neural networks and the testing of the model performance on the basis of a test set of 300 surgical incision images, the average accuracy of the four ResNet classification network models, SE-ResNet101, ResNet50, ResNet101, and SE-ResNet50, for surgical incision classification was 0.941, 0.903, 0.896, and 0.918, respectively, the precision was 0.939, 0.898, 0.868, and 0.903, respectively, and the recall rate was 0.930, 0.880, 0.850, and 0.894, respectively, with the SE-Resnet101 network model showing the highest average accuracy of 0.941 for incision feature classification. Through the combined use of deep learning technology and images of surgical incisions, problematic features of surgical incisions can be effectively identified by examining surgical incision images. It is expected that patients will eventually be able to perform self-examination of surgical incisions on smart terminals.

Convolutional neural networks applied to microtomy: Identifying the trimming-end cutting routine on paraffin-embedded tissue blocks

In the field of histopathology, the microtomy procedure yields thin sections of tissue embedded in paraffin blocks, then to be further processed for diagnostic purposes. Within microtomy, trimming is an initial but critical process in which the excess paraffin covering the tissue of interest is removed by continuous cutting routines, until the tissue is suitably exposed and ready for sectioning. Trimming is currently a time-consuming process that is manually held by technicians. In this paper, we present a method to automatize this process, by analyzing tissue block surface images resulting from each cyclic cutting routine. Two types of Convolutional Neural Networks (CNNs) were fine-tuned: one for binary segmentation, the other for multi-class classification tasks, by exploring and optimizing lightweight architectures to provide fast analytical results on cost-effective edge computing. Two sequential online conditions followed the CNNs to output the current stage of the block surface and rule if the trimming-end cutting routine was reached. We compared the results obtained through our method with the ones obtained by three skilled technicians processing 75 tissue blocks. The proposed method identified the trimming-end cutting routine approximately as accurate as the technicians did, yielding up to 90% of trimmed blocks of optimal quality and evidencing the potential of this tool in future automated trimming instruments. We deployed our method to an Edge TPU hardware accelerator to showcase its capability to provide immediate and objective results at every microtomy station applied with an ad hoc hardware, potentially guaranteeing a throughput 50% higher than manual trimming.

Wind Power Forecasting Based on WaveNet and Multitask Learning

Accurately predicting the power output of wind turbines is crucial for ensuring the reliable and efficient operation of large-scale power systems. To address the inherent limitations of physical models, statistical models, and machine learning algorithms, we propose a novel framework for wind turbine power prediction. This framework combines a special type of convolutional neural network, WaveNet, with a multigate mixture-of-experts (MMoE) architecture. The integration aims to overcome the inherent limitations by effectively capturing and utilizing complex patterns and trends in the time series data. First, the maximum information coefficient (MIC) method is applied to handle data features, and the wavelet transform technique is employed to remove noise from the data. Subsequently, WaveNet utilizes its scalable convolutional network to extract representations of wind power data and effectively capture long-range temporal information. These representations are then fed into the MMoE architecture, which treats multistep time series prediction as a set of independent yet interrelated tasks, allowing for information sharing among different tasks to prevent error accumulation and improve prediction accuracy. We conducted predictions for various forecasting horizons and compared the performance of the proposed model against several benchmark models. The experimental results confirm the strong predictive capability of the WaveNet–MMoE framework.

(433) Eros and (25143) Itokawa surface properties from reflectance spectra

Context. Our knowledge of near-Earth asteroid (NEA) composition is important for planetary research, planetary defence, and future in-space resource utilisation. Upcoming space missions, for example, Hera, M-ARGO, or missions to the asteroid (99942) Apophis, will provide us with surface-resolved NEA reflectance spectra. Neural networks are useful tools for analysing reflectance spectra and determining material composition with high precision and low processing time. Aims. We applied neural-network models on disk-resolved spectra of the Eros and Itokawa asteroids observed by the NEAR Shoemaker and Hayabusa spacecraft. With this approach, the mineral variations or intensity of space weathering can be mapped. Methods. We built and tested two types of convolutional neural networks (CNNs). The first one was trained using asteroid reflectance spectra with known taxonomy classes. The other one used silicate reflectance spectra with assigned mineral abundances and compositions. Results. The reliability of the classification model depends on the resolution of reflectance spectra. Typical F1 score and Cohen's κC values decrease from about 0.90 for high-resolution spectra to about 0.70 for low-resolution spectra. The predicted silicate composition does not strongly depend on spectrum resolution and coverage of the 2-µm band of pyroxene. The typical root mean square error is between 6 and 10 percentage points. For the Eros and Itokawa asteroids, the predicted taxonomy classes favour the S-type and the predicted surface compositions are homogeneous and correspond to the composition of L/LL and LL ordinary chondrites, respectively. On the Itokawa surface, the model identified fresh spots that were connected with craters or coarse-grain areas. Conclusions. The neural network models trained with measured spectra of asteroids and silicate samples are suitable for deriving surface silicate mineralogy with a reasonable level of accuracy. The predicted surface mineralogy is comparable to the mineralogy of returned samples measured in the laboratory. Moreover, the taxonomical predictions can point out locations of fresher areas.

Patient selection for proton therapy using Normal Tissue Complication Probability with deep learning dose prediction for oropharyngeal cancer.

In cancer care, determining the most beneficial treatment technique is a key decision affecting the patient's survival and quality of life. Patient selection for proton therapy (PT) over conventional radiotherapy (XT) currently entails comparing manually generated treatment plans, which requires time andexpertise. We developed an automatic and fast tool, AI-PROTIPP (Artificial Intelligence Predictive Radiation Oncology Treatment Indication to Photons/Protons), that assesses quantitatively the benefits of each therapeutic option. Our method uses deep learning (DL) models to directly predict the dose distributions for a given patient for both XT and PT. By using models that estimate the Normal Tissue Complication Probability (NTCP), namely the likelihood of side effects to occur for a specific patient, AI-PROTIPP can propose a treatment selection quickly andautomatically. A database of 60 patients presenting oropharyngeal cancer, obtained from the Cliniques Universitaires Saint Luc in Belgium, was used in this study. For every patient, a PT plan and an XT plan were generated. The dose distributions were used to train the two dose DL prediction models (one for each modality). The model is based on U-Net architecture, a type of convolutional neural network currently considered as the state of the art for dose prediction models. A NTCP protocol used in the Dutch model-based approach, including grades II and III xerostomia and grades II and III dysphagia, was later applied in order to perform automatic treatment selection for each patient. The networks were trained using a nested cross-validation approach with 11-folds. We set aside three patients in an outer set and each fold consists of 47 patients in training, five in validation and five for testing. This method allowed us to assess our method on 55 patients (five patients per test times the number of folds). The treatment selection based on the DL-predicted doses reached an accuracy of 87.4% for the threshold parameters set by the Health Council of the Netherlands. The selected treatment is directly linked with these threshold parameters as they express the minimal gain brought by the PT treatment for a patient to be indicated to PT. To validate the performance of AI-PROTIPP in other conditions, we modulated these thresholds, and the accuracy was above 81% for all the considered cases. The difference in average cumulative NTCP per patient of predicted and clinical dose distributions is very similar (less than 1% difference). AI-PROTIPP shows that using DL dose prediction in combination with NTCP models to select PT for patients is feasible and can help to save time by avoiding the generation of treatment plans only used for the comparison. Moreover, DL models are transferable, allowing, in the future, experience to be shared with centers that would not have PT planningexpertise.

Automatic Modulation Recognition of Radiation Source Signals Based on Data Rearrangement and the 2D FFT

It is a challenge for automatic modulation recognition (AMR) methods for radiation source signals to work in environments with low signal-to-noise ratios (SNRs). This paper proposes a modulation feature extraction method based on data rearrangement and the 2D fast Fourier transform (FFT) (DR2D), and a DenseNet feature extraction network with early fusion is constructed to recognize the extracted modulation features. First, the input signal is preprocessed by DR2D to obtain three types of joint frequency feature bins with multiple time scales. Second, the feature fusion operation is performed on the inputs of the different layers of the proposed network. Finally, feature recognition is completed in the subsequent layers. The theoretical analysis and simulation results show that DR2D is a fast and robust preprocessing method for extracting the features of modulated radiation source signals with less computational complexity. The proposed DenseNet feature extraction network with early fusion can identify the extracted modulation features with less spatial complexity than other types of convolutional neural networks (CNNs) and performs well in low-SNR environments.