Implementation Of Convolutional Neural Network Research Articles

Rapid Prediction of Multidrug-Resistant Klebsiella pneumoniae through Deep Learning Analysis of SERS Spectra.

Klebsiella pneumoniae is listed by the WHO as a priority pathogen of extreme importance that can cause serious consequences in clinical settings. Due to its increasing multidrug resistance all over the world, K. pneumoniae has the potential to cause extremely difficult-to-treat infections. Therefore, rapid and accurate identification of multidrug-resistant K. pneumoniae in clinical diagnosis is important for its prevention and infection control. However, the limitations of conventional and molecular methods significantly hindered the timely diagnosis of the pathogen. As a label-free, noninvasive, and low-cost method, surface-enhanced Raman scattering (SERS) spectroscopy has been extensively studied for its application potentials in the diagnosis of microbial pathogens. In this study, we isolated and cultured 121 K. pneumoniae strains from clinical samples with different drug resistance profiles, which included polymyxin-resistant K. pneumoniae (PRKP; n = 21), carbapenem-resistant K. pneumoniae, (CRKP; n = 50), and carbapenem-sensitive K. pneumoniae (CSKP; n = 50). For each strain, a total of 64 SERS spectra were generated for the enhancement of data reproducibility, which were then computationally analyzed via the convolutional neural network (CNN). According to the results, the deep learning model CNN plus attention mechanism could achieve a prediction accuracy as high as 99.46%, with robustness score of 5-fold cross-validation at 98.87%. Taken together, our results confirmed the accuracy and robustness of SERS spectroscopy in the prediction of drug resistance of K. pneumoniae strains with the assistance of deep learning algorithms, which successfully discriminated and predicted PRKP, CRKP, and CSKP strains. IMPORTANCE This study focuses on the simultaneous discrimination and prediction of Klebsiella pneumoniae strains with carbapenem-sensitive, carbapenem-resistant, and polymyxin-resistant phenotypes. The implementation of CNN plus an attention mechanism makes the highest prediction accuracy at 99.46%, which confirms the diagnostic potential of the combination of SERS spectroscopy with the deep learning algorithm for antibacterial susceptibility testing in clinical settings.

Exploring conductance modulation and implementation of convolutional neural network in Pt/ZnO/Al2O3/TaN memristors for brain-inspired computing

Oxide-based memristors have emerged as a promising electronic device for high-density memory and neuromorphic applications. In our study, we explored the tunable analog switching and biological synaptic functions of a Pt/ZnO/Al2O3/TaN memristive device. Using transmission electron microscopy (TEM) and x-ray photoelectron spectroscopy (XPS), we confirmed the presence of a TaOxNy interface layer at the anode contact, believed to play a critical role in resistance transitions. The memristive device showed excellent performance, including a stable and reproducible analog switching memory with a low operating voltage (μ=̶2.0/+1.7V), good cycling endurance (2 × 102), a high on/off ratio (>103), and retention up to 104 s at 85 °C. Additionally, multi-state resistances were achieved by varying the reset voltage, enabling the creation of neuromorphic synapses and high-density memories. Direct-current mode set and reset transitions showed multi-state resistance changes similar to potentiation and depression behaviors in biological synapses. Further simulations, including long-term potentiation (LTP) and long-term depression (LTD), paired pulse facilitation (PPF), and convolutional neural network (CNN) simulations for handwritten digits, showed an accuracy of 86.5%. These results indicate that the memristive device is highly suitable for use in high-density memory and brain-inspired computer systems.

Implementation of 1D convolutional neural network for improvement remote photoplethysmography measurement

<span lang="EN-US">Remote photoplethysmography (rPPG) for non-contact heart rate measurement has been widely developed and shows good development. However, motion artifact due to changes in illumination and subject movement is still the main problem. Especially when measurements are taken in real conditions. In these conditions, it will be vulnerable to rPPG signal readings with poor signal quality. So, in this paper, it is proposed to classify the signal quality using one dimensional convolutional neural network (1D CNN). The classification is carried out based on the extraction of the temporal features of the rPPG signal that has been obtained from the plane orthogonal to skin algorithm and the magnitude of the subject's movement when measured. The classification results are entered into a compensated network if the signal obtained shows moderate quality. The compensated network will provide a more accurate estimate of hr value. The test was carried out using a dataset of 10 subjects, each measured with 3 different types of illumination. In the experiments conducted, the system's performance showed an improvement compared to the POS algorithm alone. The experiment found that the mean absolute error measurement was 2.78, and the mean error was relative at 3.67%.</span>

Multi-input parallel graph neural network for semi-supervised rolling bearing fault diagnosis

Rolling bearing fault diagnosis is the key technology to ensure the reliable, efficient and sustainable operation of rotating machinery. Many fault diagnosis methods have been proposed based on vibration signal analysis from the perspective of data-driven analytics. However, these methods normally take signals of multiple sensors as a whole for feature extraction without considering the relationship among samples. This drawback leads to insufficient feature mining, thereby affecting the accuracy of fault diagnosis. Moreover, these methods need large numbers of labeled samples to achieve high diagnosis accuracy, which requires extensive human labor and is impractical in many real-world applications. To address these issues, a semi-supervised rolling bearing fault diagnosis method based on multi-input parallel graph neural network is proposed in this paper. In the proposed model, signals of multiple sensors are treated separately; thus, features will be extracted parallelly in a more sufficient way. Then, signals of each sensor are constructed into a graph based on limited-radius nearest neighbor, which will add extra relationship information to aid in fault diagnosis. In addition, with the implementation of graph convolutional neural network, the proposed method is able to achieve a more accurate diagnosis than the comparison methods in the case of few labeled data. Finally, the proposed model is evaluated on rolling bearing dataset provided by Case Western Reserve University. Compared with some classical fault diagnosis methods, the proposed model can improve the diagnosis accuracy up to more than 99% even when the proportion of training samples is only 20%.

Unconventional Integrated Photonic Accelerators for High-Throughput Convolutional Neural Networks

We provide an overview of the rapidly evolving landscape of integrated photonic neuromorphic architectures, specifically targeting the implementation of convolutional neural networks. The exploding research momentum stems from the well-known advantages of photonic circuits compared to digital electronics, and at the same time, it is driven by the massive need for cognitive image/video processing. In this context, we provide a detailed literature review on photonic cores operating as convolutional neural networks, covering either the functionality of a conventional neural network or its spiking counterpart. Moreover, we propose 2 alternative photonic approaches that refrain from simply transferring neural network concepts directly into the optical domain; instead, they focus on fusing photonic, digital electronic, and event-based bioinspired processing to optimally exploit the virtues of each scheme. These approaches can offer beyond state-of-the-art performance while relying on realistic, scalable technology. The first approach is based on a photonic integrated platform and a bioinspired spectrum-slicing technique. The photonic chip allows feature extraction through optical filtering with low power consumption and an equivalent computational efficiency of 72 femtojoules per multiply-and-accumulate operation for 5-bit precision. When combined with typical digital neural networks, an almost 5-fold reduction in the number of parameters was achieved with a minor loss of accuracy compared to established convolutional neural networks. The second approach follows a bioisomorphic route in which miniaturized spiking laser neurons and unsupervised bioinspired training are unified in a deep architecture, revealing a noise-resilient and power-efficient proposition.

Artificial intelligence-based Medicine Recognition System using faster Recurrent Convolutional Neural Networks

Recently, chronic patients are taking multiple medications incorrectly and taking the wrong medications due to similarity of drugs.It is possible that taking the improper medication can result in hazardous interactions with other medications or they will counteract the intended benefits of the medications, resulting in extra severe repercussions such as acute complications. The conventional methods are failed to provide the maximum efficiency. Therefore, this article is focused on implementation of faster recurrent convolutional neural networks (FR-CNN), which is capable of extracting the features from images. FR-CNN mainly used to analyze the patterns of the medicines and extracts the deep features. Further, classification of medicines is carried out by comparing with ground truth labels. The simulation results shows that the proposed system resulted in superior performance as compared to state of art approaches.

Implementation of CNN for Plant Identification using UAV Imagery

Plants are the world's most significant resource since they are the only natural source of oxygen. Additionally, plants are considered crucial since they are the major source of energy for humanity and have nutritional, therapeutic, and other benefits. Image identification has become more prominent in this technology-driven world, where many innovations are happening in this sphere. Image processing techniques are increasingly being used by researchers to identify plants. The capacity of Convolutional Neural Networks (CNN) to transfer weights learned with huge standard datasets to tasks with smaller collections or more particular data has improved over time. Several applications are made for image identification using deep learning, and Machine Learning (ML) algorithms. Plant image identification is a prominent part of such. The plant image dataset of about 300 images collected by mobile phone and camera from different places in the natural scenes with nine species of different plants are deployed for training. A five-layered convolution neural network (CNN) is applied for large-scale plant classification in a natural environment. The proposed work claims a higher accuracy in plant identification based on experimental data. The model achieves the utmost recognition rate of 96% NU108 dataset and UAV images of NU101 have achieved an accuracy of 97.8%.

A Conversion of Signal to Image Method for Two-Dimension Convolutional Neural Networks Implementation in Power Quality Disturbances Identification

The power quality is identified and monitored to prevent the worst effects arise on the electrical devices. These effects can be device failure, performance degradation, and replacement of some device parts. The deep convolutional neural networks (DCNNs) method can extract the complexity of image features. This method is adopted for the power quality disruption identification of the model. However, the power quality signal data is a time series. Therefore, this paper proposes an approach for the conversion of a power quality disturbance signal to an image. This research is conducted in several stages for constructing the approach proposed. Firstly, the size of a matrix is determined based on the sampling frequency values and cycle number of the signal. Secondly, a zero-cross algorithm is adopted to specify the number of signal sample points inserted into rows of the matrix. The matrix is then converted into a grayscale image. Furthermore, the resulting images are fed to the two-dimension (2D) CNNs model for the PQDs feature learning process. When the classification model is fit, then the model is tested for power quality data prediction. Finally, the model performance is evaluated by employing the confusion matrix method. The model testing result exhibits that the parameter values such as accuracy, recall, precision, and f1-score achieve at 99.81%, 98.95%, 98.84, and 98.87 %, respectively. In addition, the proposed method's performance is superior to the previous methods.Â

Implementation of convolutional neural network for predicting glaucoma from fundus images

The contributions of this paper are two-fold. First, it uses machine learning tools to detect and monitor glaucoma. Second, it provides insight into medicine, which can assist healthcare professionals in improving disease diagnostic accuracy and help to reduce the progression and degeneration of retinal ganglion cells in patients. Glaucoma is a group of eye diseases that damage the optic nerve, which can cause vision loss and blindness at any age. The main symptom of glaucoma in the early stages is high internal eye pressure. It is still unknown what causes these diseases to develop, but if not treated, they result in optic nerve atrophy. For this reason, in this paper we propose a novel deep learning system for the automatic diagnosis of glaucoma using a convolutional neural network for classification, which demonstrates improved performance and records computation time for fundus images. The results showed an accuracy of 94 % and a loss value of only 0.27. The model we have created to investigate with Keras helped us achieve good results in our training and testing process. These study results demonstrate the ability of a deep learning model to identify glaucoma from fundus images. Increasing the filter size and training the model resulted in a higher accuracy rate. A population survey that was conducted in 2019 shows that most patients with glaucoma become aware of their disease late, after the disease causes a high level of optic nerve damage and a high percentage of vision loss. Early diagnosis and detection of glaucoma using optic nerve imaging technology have gained wide clinical interest in stopping or slowing the progression of the disease, allowing the development of new algorithms to automate the diagnosis of eye diseases

Enhancing Solar Energy Forecast Using Multi-Column Convolutional Neural Network and Multipoint Time Series Approach

The rapid expansion of solar industries presents unknown technological challenges. A dedicated and suitable energy forecast is an effective solution for the daily dispatching and production of the electricity grid. The traditional forecast technique uses weather and plant parameters as the model information. Nevertheless, these are insufficient to consider problematic weather variability and the various plant characteristics in the actual field. Considering the above facts and inspired by the excellent implementation of the multi-column convolutional neural network (MCNN) in image processing, we developed a novel approach for forecasting solar energy by transforming multipoint time series (MT) into images for the MCNN to examine. We first processed the data to convert the time series solar energy into image matrices. We observed that the MCNN showed a preeminent response under a ground-based high-resolution spatial–temporal image matrix with a 0.2826% and 0.5826% RMSE for 15 min-ahead forecast under clear (CR) and cloudy (CD) conditions, respectively. Our process was performed on the MATLAB deep learning platform and tested on CR and CD solar energy conditions. The excellent execution of the suggested technique was compared with state-of-the-art deep neural network solar forecasting techniques.

Human activity recognition based on integration of multilayer information of convolutional neural network architecture

SummaryHuman activity recognition (HAR) has gained researcher's interest due to its increasing demand in automated monitoring applications. Development of efficient HAR algorithm is still an open research area due to the challenges like inter and intra‐class variations, diversity in lighting conditions, view point changes, and complex object motions. Convolutional neural network (CNN) based methods have achieved significant improvement in HAR. However, CNN implementations have drawback that it require a lot of computational resources due to the use of large number of learnable parameters. To overcome this drawback, we propose a simple and computationally efficient deep CNN architecture using multi‐layer information fusion for HAR. In this study, we explore the impact of information fusion at intermediate layers of the network, as each convolutional layer of the network hierarchically extracts information at different level of abstraction of the objects from the video frames. In this work, first we designed a simple and computationally efficient deep CNN architecture and then we introduce a feature fusion strategy to integrate the complementary information of intermediate layers to the layer of the proposed CNN architecture. The proposed architecture is fine‐tuned and trained from scratch with raw RGB data. Softmax classifier is used at the last layer of network for activity classification. Benefits of the proposed architecture over standard deep learning architectures is it's high computational efficiency and reduced requirement of computational resources. To prove the effectiveness of the proposed method, we performed several extensive experiments on publically available datasets. The experimental results of the proposed method have demonstrated its superiority over other existing state‐of‐the‐art methods.

Implementation of convolutional NEURAL network in FPGA for image recognition

Implementation of convolutional NEURAL network in FPGA for image recognition

1-ADM-CNN: A Lightweight In-field Compression Method for Seismic Data

Large-scale seismic acquisition, versatility, flexibility, automation, and scalability are the objectives of future oil and gas exploration technology. An example of emerging technology for seismic monitoring is distributed acoustic sensing (DAS). The significant amount of data produced by DAS is a challenge that necessitates the development of new technologies for its efficient handling and processing. A typical seismic survey, on the one hand, can generate hundreds of terabytes of raw seismic data per day. The demand for wireless seismic data transmission, on the other hand, remains enormous. The massive amount of data transmission from geophones to the on-site data collection center and its storage poses significant challenges. A lightweight compression procedure is required in order to reduce the data traffic and the storage size at the data center without putting an extra burden on a geophone. In this brief, an efficient implementation of a 1D convolutional neural network (CNN) together with 1-bit adaptive delta modulation is presented for in-field seismic data compression. It is worth mentioning here that the training of CNN is done offline on synthetic data set and hence, the proposed approach has potential for real-time implementation. Furthermore, no assumption on the underlying statistics of noise or the seismic signal is imposed and consequently, the proposed method is suitable for a wide range of seismic data. Furthermore, the proposed method works in the time-domain, unlike existing transform-domain methods, making it suitable for quick diagnosis of bad traces at the data center. Simulation results with real data set reveal that the proposed approach achieves a signal-to-noise ratio (SNR) of approximately 30 dB with a compression gain of 35: 1. Finally, significant superiority in terms of compression gain and reconstruction quality is demonstrated when compared to the existing methods.

Convolutional Neural Network Approach and Implementation Sensor Design for Detecting Object

This research offered sensor design for detecting animal for any applications uses convolutional neural network within JeVois smart camera. The camera is connected to communication processor through serial connection. By using 900 images for training, the sensor has been able to detect the intended animal presence and send the result to end node. The acuracy achieves 95 %, with 90.91% precision, recall 100% and F score of 95.24 %. Even dough the power consumption increases from 2.2 W to 4.08 W, the detection is potentially reduces number of the transmitter images.

An Energy-Efficient Mixed-Bit CNN Accelerator With Column Parallel Readout for ReRAM-Based In-Memory Computing

Computing-In-memory (CIM) accelerators have the characteristics of storage and computing integration, which has the potential to break through the limit of Moore’s law and the bottleneck of Von-Neumann architecture for convolutional neural networks (CNN) implementation improvement. However, the performance of CIM accelerators is still limited by conventional CNN architectures and inefficient readouts. To increase energy-efficient performance, an optimized CNN model is required and a low-power column parallel readout is necessary for edge-computing hardware. In this work, an ReRAM-based CNN accelerator is designed. Mixed-bit operations from 1 bit to 8 bits are supported by an effective bitwidth configuration scheme to implement Neural Architecture Search (NAS)-optimized layer-wise multi-bit CNNs. Besides, column-parallel readout is achieved with excellent energy-efficient performance by a variation-reduction accumulation mechanism and low-power readout circuits. Additionally, we further explore systolic data reuse in an ReRAM-based PE array. Experiments are implemented on NAS-optimized ResNet-18. Benchmarks show that the proposed ReRAM accelerator can achieve peak energy efficiency of 2490.32 TOPS/W for 1-bit operation and average energy efficiency of 479.37 TOPS/W for <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$1\sim 8$ </tex-math></inline-formula> -bit operations with evaluating NAS-optimized multi-bitwidth CNNs. When compared with the state-of-the-art works, the proposed accelerator shows at least <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$14.18{\times }$ </tex-math></inline-formula> improvement on energy efficiency.

A deep learning generative model approach for image synthesis of plant leaves.

A well-known drawback to the implementation of Convolutional Neural Networks (CNNs) for image-recognition is the intensive annotation effort for large enough training dataset, that can become prohibitive in several applications. In this study we focus on applications in the agricultural domain and we implement Deep Learning (DL) techniques for the automatic generation of meaningful synthetic images of plant leaves, which can be used as a virtually unlimited dataset to train or validate specialized CNN models or other image-recognition algorithms. Following an approach based on DL generative models, we introduce a Leaf-to-Leaf Translation (L2L) algorithm, able to produce collections of novel synthetic images in two steps: first, a residual variational autoencoder architecture is used to generate novel synthetic leaf skeletons geometry, starting from binarized skeletons obtained from real leaf images. Second, a translation via Pix2pix framework based on conditional generator adversarial networks (cGANs) reproduces the color distribution of the leaf surface, by preserving the underneath venation pattern and leaf shape. The L2L algorithm generates synthetic images of leaves with meaningful and realistic appearance, indicating that it can significantly contribute to expand a small dataset of real images. The performance was assessed qualitatively and quantitatively, by employing a DL anomaly detection strategy which quantifies the anomaly degree of synthetic leaves with respect to real samples. Finally, as an illustrative example, the proposed L2L algorithm was used for generating a set of synthetic images of healthy end diseased cucumber leaves aimed at training a CNN model for automatic detection of disease symptoms. Generative DL approaches have the potential to be a new paradigm to provide low-cost meaningful synthetic samples. Our focus was to dispose of synthetic leaves images for smart agriculture applications but, more in general, they can serve for all computer-aided applications which require the representation of vegetation. The present L2L approach represents a step towards this goal, being able to generate synthetic samples with a relevant qualitative and quantitative resemblance to real leaves.

Detection of Parthenium Weed (Parthenium hysterophorus L.) and Its Growth Stages Using Artificial Intelligence

Parthenium weed (Parthenium hysterophorus L. (Asteraceae)), native to the Americas, is in the top 100 most invasive plant species in the world. In Australia, it is an annual weed (herb/shrub) of national significance, especially in the state of Queensland where it has infested both agricultural and conservation lands, including riparian corridors. Effective control strategies for this weed (pasture management, biological control, and herbicide usage) require populations to be detected and mapped. However, the mapping is made difficult due to varying nature of the infested landscapes (e.g., uneven terrain). This paper proposes a novel method to detect and map parthenium populations in simulated pastoral environments using Red-Green-Blue (RGB) and/or hyperspectral imagery aided by artificial intelligence. Two datasets were collected in a control environment using a series of parthenium and naturally co-occurring, non-parthenium (monocot) plants. RGB images were processed with a YOLOv4 Convolutional Neural Network (CNN) implementation, achieving an overall accuracy of 95% for detection, and 86% for classification of flowering and non-flowering stages of the weed. An XGBoost classifier was used for the pixel classification of the hyperspectral dataset—achieving a classification accuracy of 99% for each parthenium weed growth stage class; all materials received a discernible colour mask. When parthenium and non-parthenium plants were artificially combined in various permutations, the pixel classification accuracy was 99% for each parthenium and non-parthenium class, again with all materials receiving an accurate and discernible colour mask. Performance metrics indicate that our proposed processing pipeline can be used in the preliminary design of parthenium weed detection strategies, and can be extended for automated processing of collected RGB and hyperspectral airborne unmanned aerial vehicle (UAV) data. The findings also demonstrate the potential for images collected in a controlled, glasshouse environment to be used in the preliminary design of invasive weed detection strategies in the field.

POD-RACING: Bulk-Bitwise to Floating-Point Compute in Racetrack Memory for Machine Learning at the Edge

Convolutional neural networks (CNNs) have become a ubiquitous algorithm with growing applications in mobile and edge settings. We describe a compute-in-memory (CIM) technique called POD-RACING using Racetrack memory (RM) to accelerate CNNs for edge systems. Using transverse read, a technique that can determine the number of “1”s in multiple adjacent domains, POD-RACING can efficiently implement multioperand bulk-bitwise and addition computations, and two-operand multiplication. We discuss how POD-RACING can implement both variable precision integer and floating point arithmetic using digital CIM. This allows both CNN inference and on-device training without expensive data movement to the cloud. Based on these functions we demonstrate the implementation of several CNNs with backpropagation using RM CIM and compare these to the state-of-the-art implementations of CNN inference and training. During training, POD-RACING improves efficiency by 2×, energy consumption by $\geq$≥27%, and increases throughput by $\geq$≥18% versus a state-of-the-art field-programmable gate array accelerator.

Efficient Hardware Acceleration of Sparsely Active Convolutional Spiking Neural Networks

Spiking neural networks (SNNs) compute in an event-based manner to achieve a more efficient computation than standard neural networks. In SNNs, neuronal outputs are not encoded as real-valued activations but as sequences of binary spikes. The motivation of using SNNs over conventional neural networks is rooted in the special computational aspects of spike-based processing, especially the high degree of sparsity of spikes. Well-established implementations of convolutional neural networks (CNNs) feature large spatial arrays of processing elements (PEs) that remain highly underutilized in the face of activation sparsity. We propose a novel architecture optimized for the processing of convolutional SNNs (CSNNs) featuring a high degree of sparsity. The proposed architecture consists of an array of PEs of the size of the kernel of a convolution and an intelligent spike queue that provides a high PE utilization. A constant flow of spikes is ensured by compressing the feature maps into queues that can then be processed spike-by-spike. This compression is performed at run-time, leading to a self-timed schedule. This allows the processing time to scale with the number of spikes. Also, a novel memory organization scheme is introduced to efficiently store and retrieve the membrane potentials of the individual neurons using multiple small parallel on-chip RAMs. Each RAM is hardwired to its PE, reducing switching circuitry. We implemented the proposed architecture on an FPGA and achieved a significant speedup compared to previously proposed SNN implementations (~10 times) while needing less hardware resources and maintaining a higher energy efficiency (~15 times).

CrackW-Net: A Novel Pavement Crack Image Segmentation Convolutional Neural Network

Image-based intelligent detection of road cracks with high accuracy and efficiency is vital to the overall condition assessment of the pavement. However, significant problems of continuous cracks interruption and background discrete noise misidentification are frequently observed in current semantic segmentation of pavement cracks, which mainly caused by traditional segmentation convolutional neural networks. This paper proposes a skip-level round-trip sampling block structure with the implementation of convolutional neural networks, thereby constructed a novel pixel level semantic segmentation network called CrackW-Net. After that, two datasets, including the widely recognized Crack500 dataset and a self-built dataset, were used to train two versions CrackW-Net, FCN, U-Net and ResU-Net. Meanwhile, comparative experiments are conducted among all these network models for crack detection. Results show that CrackW-Net without residual block performs the best in the task of pavement crack segmentation.