Lightweight Convolutional Neural Network Architecture Research Articles

CodnNet: A lightweight CNN architecture for detection of COVID-19 infection

The application of Convolutional Neural Network (CNN) on the detection of COVID-19 infection has yielded favorable results. However, with excessive model parameters, the CNN detection of COVID-19 is low in recall, highly complex in computation. In this paper, a novel lightweight CNN model, CodnNet is proposed for quick detection of COVID-19 infection. CodnNet builds a more effective dense connections based on DenseNet network to make features highly reusable and enhances interactivity of local and global features. It also uses depthwise separable convolution with large convolution kernels instead of traditional convolution to improve the range of receptive field and enhances classification performance while reducing model complexity. The 5-Fold cross validation results on Kaggle’s COVID-19 Dataset showed that CodnNet has an average precision of 97.9%, recall of 97.4%, F1score of 97.7%, accuracy of 98.5%, mAP of 99.3%, and mAUC of 99.7%. Compared to the typical CNNs, CodnNet with fewer parameters and lower computational complexity has achieved better classification accuracy and generalization performance. Therefore, the CodnNet model provides a good reference for quick detection of COVID-19 infection.

A deep learning-based framework for detecting COVID-19 patients using chest X-rays.

Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) has caused outbreaks of new coronavirus disease (COVID-19) around the world. Rapid and accurate detection of COVID-19 coronavirus is an important step in limiting the spread of the COVID-19 epidemic. To solve this problem, radiography techniques (such as chest X-rays and computed tomography (CT)) can play an important role in the early prediction of COVID-19 patients, which will help to treat patients in a timely manner. We aimed to quickly develop a highly efficient lightweight CNN architecture for detecting COVID-19-infected patients. The purpose of this paper is to propose a robust deep learning-based system for reliably detecting COVID-19 from chest X-ray images. First, we evaluate the performance of various pre-trained deep learning models (InceptionV3, Xception, MobileNetV2, NasNet and DenseNet201) recently proposed for medical image classification. Second, a lightweight shallow convolutional neural network (CNN) architecture is proposed for classifying X-ray images of a patient with a low false-negative rate. The data set used in this work contains 2,541 chest X-rays from two different public databases, which have confirmed COVID-19 positive and healthy cases. The performance of the proposed model is compared with the performance of pre-trained deep learning models. The results show that the proposed shallow CNN provides a maximum accuracy of 99.68% and more importantly sensitivity, specificity and AUC of 99.66%, 99.70% and 99.98%. The proposed model has fewer parameters and low complexity compared to other deep learning models. The experimental results of our proposed method show that it is superior to the existing state-of-the-art methods. We believe that this model can help healthcare professionals to treat COVID-19 patients through improved and faster patient screening.

A light-weight convolutional Neural Network Architecture for classification of COVID-19 chest X-Ray images.

The COVID-19 pandemic has opened numerous challenges for scientists to use massive data to develop an automatic diagnostic tool for COVID-19. Since the outbreak in January 2020, COVID-19 has caused a substantial destructive impact on society and human life. Numerous studies have been conducted in search of a suitable solution to test COVID-19. Artificial intelligence (AI) based research is not behind in this race, and many AI-based models have been proposed. This paper proposes a lightweight convolutional neural network (CNN) model to classify COVID and Non_COVID patients by analyzing the hidden features in the X-Ray images. The model has been evaluated with different standard metrics to prove the reliability of the model. The model obtained 98.78%, 93.22%, and 92.7% accuracy in the training, validation, and testing phases. In addition, the model achieved 0.964 scores in the Area Under Curve (AUC) metric. We compared the model with four state-of-art pre-trained models (VGG16, InceptionV3, DenseNet121, and EfficientNetB6). The evaluation results demonstrate that the proposed CNN model is a candidate for an automatic diagnostic tool for the classification of COVID-19 patients using chest X-ray images. This research proposes a technique to classify COVID-19 patients and does not claim any medical diagnosis accuracy.

Driver Behaviors Recognizer Based on Light-Weight Convolutional Neural Network Architecture and Attention Mechanism

Driving is a set of behaviors that need high concentration. Sometimes these behaviors are dominated by other acts such as smoking, eating, drinking, talking, phone calls, adjusting the radio, or drowsiness. These are also the main causes of current traffic accidents. Therefore, developing applications to warn drivers in advance is essential. This research introduces a light-weight convolutional neural network architecture to recognize driver behaviors, helping the warning system to provide accurate information and to minimize traffic collisions. This network is a combination of feature extraction and classifier modules. The feature extraction module uses the advantages of the standard convolution layers, depthwise separable convolution layers, average pooling layers, and proposed adaptive connections to extract the feature maps. The benefit of the convolution block attention module is deployed in the feature extraction module that guides the network in learning the salient features. The classifier module is comprised of a global average pooling and softmax layer to calculate the probability of each class. The overall design optimizes the network parameters and maintains classification accuracy. The entire network is trained and evaluated on three benchmark datasets: the State Farm Distracted Driver Detection, the American University in Cairo version 1, and the American University in Cairo version 2. As a result, the accuracies on overall classes (ten classes) are 99.95&#x0025;, 95.57&#x0025;, and 99.61&#x0025;, respectively. Also, several video tests with VGA (Video Graphics Array), HD (High Definition), and FHD (Full High Definition) resolution were conducted, and they can be seen at <uri>https://bit.ly/3GY2iJl</uri>.

Patent ductus arteriosus (PDA) detection in echocardiograms using deep learning

BackgroundPatent ductus arteriosus (PDA) is a common form of congenital heart disease, especially in preterm infants. A PDA can be associated with prolonged ventilator dependence and increased risk of severe lung disease, necrotizing enterocolitis, impaired renal function, intraventricular hemorrhage, and death. The problem of caring for neonates with a PDA is difficult, and the use of artificial intelligence (AI) to aid in PDA detection can assist in its management. MethodsA clinical database was searched for echocardiograms performed in the Neonatal Intensive Care Unit (NICU) at the Children's Hospital of Orange County (CHOC) from 2017 to 21. A total of 461 de-identified echocardiogram video clips across 300 patients from CHOC were analyzed. Our goal was to explore the efficacy of the Convolutional Neural Network (CNN) for PDA detection in echocardiogram video clips for eventual clinical deployment on an edge-based device. To this end, we used the light-weight MobileNet-V2 CNN architecture for training and testing. Of the 461 echocardiogram video clips, 316 were used for training, 74 for validation, and 72 for testing. Video frames were extracted from each clip and processed by the CNN. The CNN treated the frames as independent images and performed binary (Normal vs PDA) classification on each video clip. Results and conclusionOf the 461 echocardiogram video clips analyzed, 272 contained an identifiable PDA and 190 were considered normal. Our CNN algorithm achieved notable results for identifying the presence of a PDA, with 0.88 Area Under Curve (AUC), 0.84 Positive Predictive Value (PPV), 0.80 Negative Predictive Value (NPV), 0.76 Sensitivity, and 0.87 Specificity on the test data. Results indicate that diagnosis of PDA within an edge-based AI framework is feasible. Future work will involve augmenting the echocardiogram dataset, expanding the analysis to include PDA classification based on size and hemodynamic significance, and exploring additional algorithmic approaches.

Light-SDNet: A Lightweight CNN Architecture for Ship Detection

Ship detection plays a vital role in monitoring and managing maritime safety. Most recently proposed learning-based object detection methods have achieved marked progress in detection accuracy, but the size of these models is too large to be applied to mobile devices with limited resources. Although some compact models have been presented in the previous study, they achieve unsatisfactory results in ship detection, especially under extreme weather conditions. To address these challenges, this article presents a lightweight convolutional neural network (CNN) called Light-SDNet to perform an end-to-end ship detection under different weather conditions. In the proposed model, we introduce the improved CA-Ghost, C3Ghost, and DepthWise Convolution (DWConv) into the You Only Look Once version 5 (YOLOv5) to reduce the number of model parameters, while remaining its powerful feature expression ability. We use parallel attention to highlight the features that contribute to the ship detection in the marine surveillance. To enhance the adaptability of the proposed model, a hybrid training strategy with generating synthetically-degraded images is proposed to augment the volume and diversity of the original datasets. The proposed strategy enables Light-SDNet to improve the ship detection results under severe weather conditions such as haze, rain, and low illumination. We compare Light-SDNet with other competitive approaches on a large-scaled ship dataset called SeaShips. We show that Light-SDNet achieves a better balance between the detection accuracy and the model complexity. The ship detection results on degraded marine images have proven the superior performance of the proposed model in terms of detection accuracy, robustness and efficiency.

A Lightweight CNN Architecture for Automatic Modulation Classification

Automatic modulation classification (AMC) algorithms based on deep learning (DL) have been widely studied in the past decade, showing significant performance advantage compared to traditional ones. However, the existing DL methods generally behave worse in computational complexity. For this, this paper proposes a lightweight convolutional neural network (CNN) for AMC task, where we design a depthwise separable convolution (DSC) residual architecture for feature extraction to prevent the vanishing gradient problem and lighten the computational burden. Besides that, in order to further reduce model complexity, global depthwise convolution (GDWConv) is adopted for feature reconstruction after the last (non-global) convolutional layer. Compared to recent works, the experimental results show that the proposed network can save approximately 70~98% model parameters and 30~99% inference time on two well-known benchmarks.

Light-Convolution Dense Selection U-Net (LDS U-Net) for Ultrasound Lateral Bony Feature Segmentation

Scoliosis is a widespread medical condition where the spine becomes severely deformed and bends over time. It mostly affects young adults and may have a permanent impact on them. A periodic assessment, using a suitable modality, is necessary for its early detection. Conventionally, the usually employed modalities include X-ray and MRI, which employ ionising radiation and are expensive. Hence, a non-radiating 3D ultrasound imaging technique has been developed as a safe and economic alternative. However, ultrasound produces low-contrast images that are full of speckle noise, and skilled intervention is necessary for their processing. Given the prevalent occurrence of scoliosis and the limitations of scalability of human expert interventions, an automatic, fast, and low-computation assessment technique is being developed for mass scoliosis diagnosis. In this paper, a novel hybridized light-weight convolutional neural network architecture is presented for automatic lateral bony feature identification, which can help to develop a fully-fledged automatic scoliosis detection system. The proposed architecture, Light-convolution Dense Selection U-Net (LDS U-Net), can accurately segment ultrasound spine lateral bony features, from noisy images, thanks to its capabilities of smartly selecting only the useful information and extracting rich deep layer features from the input image. The proposed model is tested using a dataset of 109 spine ultrasound images. The segmentation result of the proposed network is compared with basic U-Net, Attention U-Net, and MultiResUNet using various popular segmentation indices. The results show that LDS U-Net provides a better segmentation performance compared to the other models. Additionally, LDS U-Net requires a smaller number of parameters and less memory, making it suitable for a large-batch screening process of scoliosis without a high computational requirement.

Lightweight Density Map Architecture for UAVs Safe Landing in Crowded Areas

In the context of crowd counting, most of the works have focused on improving the accuracy without regard to the performance leading to algorithms that are not suitable for embedded applications. In this paper, we propose a lightweight convolutional neural network architecture to perform crowd detection and counting using fewer computer resources without a significant loss on count accuracy. The architecture was trained using the Bayes loss function to further improve its accuracy and then pruned to further reduce the computational resources used. The proposed architecture was tested over the USF-QNRF achieving a competitive Mean Average Error of 154.07 and a superior Mean Square Error of 241.77 while maintaining a competitive number of parameters of 0.067 Million. The obtained results suggest that the Bayes loss can be used with other architectures to further improve them and also the last convolutional layer provides no significant information and even encourage over-fitting at training.

Lightweight Deep Learning Model in Mobile-Edge Computing for Radar-Based Human Activity Recognition

Radar-based human activity recognition (HAR) has great potential in many fields, such as surveillance, smart homes, and human-computer interaction. Complex deep neural networks have brought significant improvement in classification performance but also a surge of computational cost and the number of parameters, which makes it challenging to deploy in mobile devices. However, the existing studies in this area mainly focus on improving the classification accuracy. In this article, we propose an extremely efficient convolutional neural network (CNN) architecture named Mobile-RadarNet, which is specially designed for human activity classification based on micro-Doppler signatures. The new architecture exploits 1-D depthwise convolutions and pointwise convolutions to build lightweight CNN architecture. The experiments on a seven-class human activity data set demonstrate that the proposed Mobile-RadarNet can achieve high classification accuracy meanwhile to keep the computational complexity at an extremely low level, and thus has great potential to be deployed in the mobile devices.

A Hybrid Single Image Super-Resolution Technique Using Fractal Interpolation and Convolutional Neural Network

Advent of convolutional neural network (CNN) in the field of single image super-resolution (SISR) has shown immense improvement in the process of high resolution (HR) image generation. It involves an end-to-end mathematical mapping through non-linear feature extraction between low resolution (LR) and HR image. Performance of CNN can be improved by increasing depth of the architecture, which generally results in higher computational cost and running time. Also, the performance of CNN can be improved by providing more appropriate input. Presently, in most of the cases input image to a CNN is a LR image that is bi-cubically interpolated to the desired size of HR image. However, bicubic interpolation results into detail smoothing of the image. Therefore, in this paper, a hybrid of CNN and fractal interpolation based SISR algorithm is proposed for reconstruction of HR image. Here, a three layered light-weight CNN architecture is utilize which is capable of producing comparable performance with the traditional SISR techniques and fractal interpolation helps in better preservation of structural and textural properties of the HR image. Experimental results are provided to prove the efficacy of the algorithm proposed in the paper.

A Lightweight Convolutional Neural Network Architecture Applied for Bone Metastasis Classification in Nuclear Medicine: A Case Study on Prostate Cancer Patients.

Bone metastasis is among the most frequent in diseases to patients suffering from metastatic cancer, such as breast or prostate cancer. A popular diagnostic method is bone scintigraphy where the whole body of the patient is scanned. However, hot spots that are presented in the scanned image can be misleading, making the accurate and reliable diagnosis of bone metastasis a challenge. Artificial intelligence can play a crucial role as a decision support tool to alleviate the burden of generating manual annotations on images and therefore prevent oversights by medical experts. So far, several state-of-the-art convolutional neural networks (CNN) have been employed to address bone metastasis diagnosis as a binary or multiclass classification problem achieving adequate accuracy (higher than 90%). However, due to their increased complexity (number of layers and free parameters), these networks are severely dependent on the number of available training images that are typically limited within the medical domain. Our study was dedicated to the use of a new deep learning architecture that overcomes the computational burden by using a convolutional neural network with a significantly lower number of floating-point operations (FLOPs) and free parameters. The proposed lightweight look-behind fully convolutional neural network was implemented and compared with several well-known powerful CNNs, such as ResNet50, VGG16, Inception V3, Xception, and MobileNet on an imaging dataset of moderate size (778 images from male subjects with prostate cancer). The results prove the superiority of the proposed methodology over the current state-of-the-art on identifying bone metastasis. The proposed methodology demonstrates a unique potential to revolutionize image-based diagnostics enabling new possibilities for enhanced cancer metastasis monitoring and treatment.

Abstract 17189: Explaining the Deep Learning Black Box by Identifying Segments of the Single-Lead ECG Signal Used for Detection of Atrial Fibrillation

Introduction: Deep learning (DL) has proved effective for automatic identification of atrial fibrillation (AF) using single-lead ECG. Adoption and trust of DL by clinicians is limited by its black box nature. Hypothesis: Post hoc explanations can elucidate what part of ECG signal is used by the black box DL algorithm, quantifying the importance of clinically relevant features in the classification decision. Making DL decision process transparent will help its integration into clinical practice. Methods: 8,528 single-lead ECG recordings collected using AliveCor devices (PhysioNet) were used. Each signal was labeled as normal sinus rhythm, AF, other arrhythmia or noise. DL automatic classification involves a lightweight convolutional neural network architecture - MobileNet - whose performance is analyzed with an explanation method for DL. Results: Each RR interval is divided into 8 equal segments, where segment 1 follows each R peak, 4 and 5 correspond to the isoelectric baseline, and 7 to the P wave. The explanation method substitutes one of these segments with a straight line, and the corresponding change in sensitivity highlights its importance for the DL algorithm decision. MobileNet achieved a sensitivity of 92.5% to identify AF (9.4% of ECGs were in AF). Sensitivity increases by 2.5% when Segment 7 is removed, indicating that the absence of P wave leads the network to classify more frequently samples as AF.(Figure) When Segments 4 and 5 are removed, the sensitivity decreases by 2.5% and 5.0%, and by 26.7% when removed together. When all RR intervals are normalized to the same value (RR in the Figure), sensitivity for AF drops by 78.3%, showing that RR intervals are key for AF detection by DL algorithm. Conclusions: Post hoc explanations for AF detection by DL from single-lead ECG show the importance of common morphological features used for classifying AF. These methods can be used to understand the decision-making process of DL and motivate its clinical adoption.

CubeSatNet: Ultralight Convolutional Neural Network designed for on-orbit binary image classification on a 1U CubeSat

A 1U CubeSat has severe limitations on size, power and downlink capabilities. Transferring images taken by imaging payload from orbit can be a challenging. To save time and effort in downlinking and post-processing, a mechanism must be in place to sort quality data before transmitting it to the ground station. This paper presents an innovative approach combining a novel CubeSat image dataset and a lightweight Convolutional Neural Network architecture for automatically selecting images for downlink on a 1U CubeSat. Coined as CubeSatNet, the neural network is trained on 60,000 augmented images, is tiny enough to run on an ARM Cortex MCU and is tested on on-orbit data from Kyushu Institute of Technology’s BIRDS-3 CubeSats with an accuracy of 90% and F1 score of 0.92. CubeSatNet outperformed SVM, DBN and AE trained to classify CubeSat images. If implemented, the CNN could cut down operation time by about 2/3 while significantly improving the quality of received data.

EmergencyNet: Efficient Aerial Image Classification for Drone-Based Emergency Monitoring Using Atrous Convolutional Feature Fusion

Deep learning-based algorithms can provide state-of-the-art accuracy for remote sensing technologies such as unmanned aerial vehicles (UAVs)/drones, potentially enhancing their remote sensing capabilities for many emergency response and disaster management applications. In particular, UAVs equipped with camera sensors can operating in remote and difficult to access disaster-stricken areas, analyze the image and alert in the presence of various calamities such as collapsed buildings, flood, or fire in order to faster mitigate their effects on the environment and on human population. However, the integration of deep learning introduces heavy computational requirements, preventing the deployment of such deep neural networks in many scenarios that impose low-latency constraints on inference, in order to make mission-critical decisions in real time. To this end, this article focuses on the efficient aerial image classification from on-board a UAV for emergency response/monitoring applications. Specifically, a dedicated Aerial Image Database for Emergency Response applications is introduced and a comparative analysis of existing approaches is performed. Through this analysis a lightweight convolutional neural network architecture is proposed, referred to as EmergencyNet, based on atrous convolutions to process multiresolution features and capable of running efficiently on low-power embedded platforms achieving upto 20x higher performance compared to existing models with minimal memory requirements with less than 1% accuracy drop compared to state-of-the-art models.

The Deep Convolutional Neural Network for NOx Emission Prediction of a Coal-Fired Boiler

This paper presents a methodology for predicting NOx emissions of a coal-fired boiler by using real operation data, coal properties and CNN (Convolutional Neural Network). Two building blocks are carefully designed following the practical guidelines for the light weight CNN architecture design. Furthermore, the building blocks are used to develop the deep CNN-based model for NOx prediction. A comprehensive comparison among different prediction models based on DL (Deep Learning) shows that the proposed deep CNN-based prediction model outperforms other prediction models in terms of RMSE (Root Mean Square Error) criteria. The results indicate that the developed deep CNN-based prediction model has more excellent accuracy and better numerical stability. Besides, the architecture design of the DL-based prediction model has a significant impact on the performance of the prediction model.

IIRNet: A lightweight deep neural network using intensely inverted residuals for image recognition

Deep neural networks have achieved great success in many tasks of pattern recognition. However, large model size and high cost in computation limit their applications in resource-limited systems. In this paper, our focus is to design a lightweight and efficient convolutional neural network architecture by directly training the compact network for image recognition. To achieve a good balance among classification accuracy, model size, and computation complexity, we propose a lightweight convolutional neural network architecture named IIRNet for resource-limited systems. The new architecture is built based on Intensely Inverted Residual block (IIR block) to decrease the redundancy of the convolutional blocks. By utilizing two new operations, intensely inverted residual and multi-scale low-redundancy convolutions, IIR block greatly reduces its model size and computational costs while matches the classification accuracy of the state-of-the-art networks. Experiments on CIFAR-10, CIFAR-100, and ImageNet datasets demonstrate the superior performance of IIRNet on the trade-offs among classification accuracy, computation complexity, and model size, compared to the mainstream compact network architectures.

Real-Time Smile Detection using Deep Learning

Real-time smile detection from facial images is useful in many real world applications such as automatic photo capturing in mobile phone cameras or interactive distance learning. In this paper, we study different architectures of object detection deep networks for solving real-time smile detection problem. We then propose a combination of a lightweight convolutional neural network architecture (BKNet) with an efficient object detection framework (RetinaNet). The evaluation on the two datasets (GENKI-4K, UCF Selfie) with a mid-range hardware device (GTX TITAN Black) show that our proposed method helps in improving both accuracy and inference time of the original RetinaNet to reach real-time performance. In comparison with the state-of-the-art object detection framework (YOLO), our method has higher inference time, but still reaches real-time performance and obtains higher accuracy of smile detection on both experimented datasets.