Lightweight Deep Learning Model Research Articles

Detection of dental periapical lesions using retinex based image enhancement and lightweight deep learning model

Dental periapical lesions, commonly associated with inflammation around the tooth apex, pose a significant challenge in early diagnosis and treatment. This study introduces a novel approach for the detection of dental periapical lesions through the integration of Retinex-based image enhancement techniques and a lightweight deep learning model. The Retinex algorithm is employed to enhance the radiographic images, addressing issues related to inconsistent illumination and contrast. Subsequently, a tailored lightweight deep learning model is designed to efficiently extract relevant features from the enhanced images. Present methodology leverages a dataset of dental radiographs to train and evaluate the deep learning model, incorporating a diverse range of periapical lesion cases. The model is optimized for computational efficiency while maintaining high accuracy, making it suitable for deployment in resource-constrained environment. To enhance precision in lesion detection, the U-Net segmentation technique has been incorporated, providing a sophisticated approach to delineate and analyze specific areas of interest within the radiographic images. This addition further refines our diagnostic framework, contributing to the robustness of lesion identification. Experimental results demonstrate the effectiveness of the Retinex-based image enhancement in improving the visibility of periapical lesions. The lightweight deep learning model exhibits promising performance in accurately detecting and classifying dental periapical lesions, showcasing its potential for early and efficient diagnosis. The results obtained from the present model are compared with those from Convolutional Neural Network (CNN) as well as with diagnosis of expert practitioners and the model is observed to perform very well. The study contributes to the advancement of computer-aided diagnostic tools in dentistry, offering a scalable and accessible solution for the identification of dental periapical lesions through the fusion of image enhancement and lightweight deep learning techniques.

An Improved Lightweight Deep Learning Model and Implementation for Track Fastener Defect Detection with Unmanned Aerial Vehicles

Track fastener defect detection is an essential component in ensuring railway safety operations. Traditional manual inspection methods no longer meet the requirements of modern railways. The use of deep learning image processing techniques for classifying and recognizing abnormal fasteners is faster, more accurate, and more intelligent. With the widespread use of unmanned aerial vehicles (UAVs), conducting railway inspections using lightweight, low-power devices carried by UAVs has become a future trend. In this paper, we address the characteristics of track fastener detection tasks by improving the YOLOv4-tiny object detection model. We improved the model to output single-scale features and used the K-means++ algorithm to cluster the dataset, obtaining anchor boxes that were better suited to the dataset. Finally, we developed the FPGA platform and deployed the transformed model on this platform. The experimental results demonstrated that the improved model achieved an mAP of 95.1% and a speed of 295.9 FPS on the FPGA, surpassing the performance of existing object detection models. Moreover, the lightweight and low-powered FPGA platform meets the requirements for UAV deployment.

A lightweight deep learning approach for detecting electrocardiographic lead misplacement

Objective. Electrocardiographic (ECG) lead misplacement can result in distorted waveforms and amplitudes, significantly impacting accurate interpretation. Although lead misplacement is a relatively low-probability event, with an incidence ranging from 0.4% to 4%, the large number of ECG records in clinical practice necessitates the development of an effective detection method. This paper aimed to address this gap by presenting a novel lead misplacement detection method based on deep learning models. Approach. We developed two novel lightweight deep learning model for limb and chest lead misplacement detection, respectively. For limb lead misplacement detection, two limb leads and V6 were used as inputs, while for chest lead misplacement detection, six chest leads were used as inputs. Our models were trained and validated using the Chapman database, with an 8:2 train-validation split, and evaluated on the PTB-XL, PTB, and LUDB databases. Additionally, we examined the model interpretability on the LUDB databases. Limb lead misplacement simulations were performed using mathematical transformations, while chest lead misplacement scenarios were simulated by interchanging pairs of leads. The detection performance was assessed using metrics such as accuracy, precision, sensitivity, specificity, and Macro F1-score. Main results. Our experiments simulated three scenarios of limb lead misplacement and nine scenarios of chest lead misplacement. The proposed two models achieved Macro F1-scores ranging from 93.42% to 99.61% on two heterogeneous test sets, demonstrating their effectiveness in accurately detecting lead misplacement across various arrhythmias. Significance. The significance of this study lies in providing a reliable open-source algorithm for lead misplacement detection in ECG recordings. The source code is available at https://github.com/wjcai/ECG_lead_check.

RTSeg-net: A lightweight network for real-time segmentation of fetal head and pubic symphysis from intrapartum ultrasound images

The segmentation of the fetal head (FH) and pubic symphysis (PS) from intrapartum ultrasound images plays a pivotal role in monitoring labor progression and informing crucial clinical decisions. Achieving real-time segmentation with high accuracy on systems with limited hardware capabilities presents significant challenges. To address these challenges, we propose the real-time segmentation network (RTSeg-Net), a groundbreaking lightweight deep learning model that incorporates innovative distribution shifting convolutional blocks, tokenized multilayer perceptron blocks, and efficient feature fusion blocks. Designed for optimal computational efficiency, RTSeg-Net minimizes resource demand while significantly enhancing segmentation performance. Our comprehensive evaluation on two distinct intrapartum ultrasound image datasets reveals that RTSeg-Net achieves segmentation accuracy on par with more complex state-of-the-art networks, utilizing merely 1.86 M parameters—just 6 % of their hyperparameters—and operating seven times faster, achieving a remarkable rate of 31.13 frames per second on a Jetson Nano, a device known for its limited computing capacity. These achievements underscore RTSeg-Net's potential to provide accurate, real-time segmentation on low-power devices, broadening the scope for its application across various stages of labor. By facilitating real-time, accurate ultrasound image analysis on portable, low-cost devices, RTSeg-Net promises to revolutionize intrapartum monitoring, making sophisticated diagnostic tools accessible to a wider range of healthcare settings.

Enhanced fingerprint pattern classification: Integrating attention modules with lightweight deep learning models

AbstractLarge fingerprint databases can make the automated search process tedious and time‐consuming. Fingerprint pattern classification is a significant step in the identification system's complexity in terms of time and speed. Although several fingerprint algorithms have been developed for classification tasks, further improvements in performance and efficiency are still required. Most of the fingerprint algorithms use deep learning techniques. However, some deep learning techniques can be resource‐intensive and computationally expensive, while others can disregard the spatial relationships between the features used in classifying fingerprint patterns. This study proposes using lightweight deep learning models (i.e., MobileNet and EfficientNet‐B0) integrated with attention modules to classify fingerprint patterns. The two lightweight models are modified, yielding MobileNet+ and EfficientNet‐B0+ models. The lightweight deep learning models can help achieve optimal performance and reduce computational complexity. The attention modules focus on distinctive features for classification. Our proposed approach integrates four attention modules for fingerprint pattern classification into two lightweight deep learning models, that is, MobileNet+ and EfficientNet‐B0+. To evaluate our approach, we use two publicly available fingerprint datasets, that is, the NIST special database 301 dataset and the LivDet dataset. The evaluation results show that the EfficientNet‐B0+ model achieves the highest classification accuracy of 97% with only 854,086 training parameters. As a conclusion, we consider the training parameters small enough for the EfficientNet‐B0+ model to be deployed on low‐resource devices.

Mapping large-scale pine wilt disease trees with a lightweight deep-learning model and very high-resolution UAV images

ABSTRACT Pine wilt disease (PWD), caused by pine wood nematodes, has brought a great loss in ecology and economy all over the world. In China, the forest health status is also significantly affected by PWD since 1980, especially in coniferous forests and mixed forest regions. The PWD spreads very fast and can cause a healthy pine wood tree to die within a very short time. An effective way to protect other healthy pine wood trees is to discover PWD trees early. Using unmanned aerial vehicle (UAV) images can help people discover the PWD tree quickly and accurately, and a few automatical methods have been developed to monitor PWD trees including the deep learning methods. Because of the robust spatial-temporal transferability, deep learning methods have become the mainstream algorithms to monitor PWD trees. As we know, the training dataset is the most important material to train a deep learning model. However, there is still a lack of a PWD segmentation dataset so far. To fill this gap, in this paper, we have generated the first open-sourced PWD segmentation dataset based on very high-resolution UAV images to help the community conduct PWD monitor research conveniently. This dataset has 994 training samples, and each sample has visible bands with 512 × 512 pixels, and the spatial resolution of this dataset is 0.05 m. In order to train an advanced segmentation model, we have designed a lightweight deep-learning model for mobile devices or edge devices in this manuscript, named MobileSeg. The main feature of MobileSeg is its decoupling, which uses the re-parameterization technology to improve the model performance. Finally, a large-scale real-world scenario experiment with very high-resolution UAV images was utilized to validate the performance of MobileSeg, the experiment result indicated that MobileSeg has achieved the best performance compared with the recent lightweight segmentation models, and the experiment also proved the effectiveness of the proposed PWD dataset.

Research on non-intrusive load identification method based on multi-feature fusion with improved shufflenetv2

Non-intrusive load monitoring is a significant advancement in energy conservation and smart electricity usage as it enables the identification of load status and equipment type without the need for extensive sensing instruments. As one of the important steps of non-intrusive load monitoring, non-intrusive load identification plays a crucial role in the correct identification of electrical appliances. In this paper, a novel approach is presented to amalgamate multiple appliance features into a new identification feature, which is then used for appliance recognition through an improved lightweight deep learning model. This addresses the issue of low accuracy when using a single feature for appliance identification. The core concept involves initially capturing the current data of an appliance during steady state operation and fusing the image features of the appliance encoded by Gramian summation angular field, Gramian difference angular field, and Markov transition field into new recognition features for each appliance using average weighting. Subsequently, a shufflenetv2 lightweight deep learning model based on the squeeze -and-excitation module is used to mine the constructed load feature information for the load classification task. The method is then experimentally validated on the Self-test Dataset, PLAID, and WHITED dataset, resulting in recognition accuracies of 100%, 98.214%, and 99.745%, respectively. These results demonstrate the effectiveness of the proposed method in significantly improving recognition performance compared to the original approach.

LW-YOLO: Lightweight Deep Learning Model for Fast and Precise Defect Detection in Printed Circuit Boards

Printed circuit board (PCB) manufacturing processes are becoming increasingly complex, where even minor defects can impair product performance and yield rates. Precisely identifying PCB defects is critical but remains challenging. Traditional PCB defect detection methods, such as visual inspection and automated technologies, have limitations. While defects can be readily identified based on symmetry, the operational aspect proves to be quite challenging. Deep learning has shown promise in defect detection; however, current deep learning models for PCB defect detection still face issues like large model size, slow detection speed, and suboptimal accuracy. This paper proposes a lightweight YOLOv8 (You Only Look Once version 8)-based model called LW-YOLO (Lightweight You Only Look Once) to address these limitations. Specifically, LW-YOLO incorporates a bidirectional feature pyramid network for multiscale feature fusion, a Partial Convolution module to reduce redundant calculations, and a Minimum Point Distance Intersection over Union loss function to simplify optimization and improve accuracy. Based on the experimental data, LW-YOLO achieved an mAP0.5 of 96.4%, which is 2.2 percentage points higher than YOLOv8; the precision reached 97.1%, surpassing YOLOv8 by 1.7 percentage points; and at the same time, LW-YOLO achieved an FPS of 141.5. The proposed strategies effectively enhance efficiency and accuracy for deep-learning-based PCB defect detection.

A lightweight deep learning model for acute myeloid leukemia-related blast cell identification

A lightweight deep learning model for acute myeloid leukemia-related blast cell identification

Celiac disease diagnosis from endoscopic images based on multi-scale adaptive hybrid architecture model

Objective. Celiac disease (CD) has emerged as a significant global public health concern, exhibiting an estimated worldwide prevalence of approximately 1%. However, existing research pertaining to domestic occurrences of CD is confined mainly to case reports and limited case analyses. Furthermore, there is a substantial population of undiagnosed patients in the Xinjiang region. This study endeavors to create a novel, high-performance, lightweight deep learning model utilizing endoscopic images from CD patients in Xinjiang as a dataset, with the intention of enhancing the accuracy of CD diagnosis. Approach. In this study, we propose a novel CNN-Transformer hybrid architecture for deep learning, tailored to the diagnosis of CD using endoscopic images. Within this architecture, a multi-scale spatial adaptive selective kernel convolution feature attention module demonstrates remarkable efficacy in diagnosing CD. Within this module, we dynamically capture salient features within the local channel feature map that correspond to distinct manifestations of endoscopic image lesions in the CD-affected areas such as the duodenal bulb, duodenal descending segment, and terminal ileum. This process serves to extract and fortify the spatial information specific to different lesions. This strategic approach facilitates not only the extraction of diverse lesion characteristics but also the attentive consideration of their spatial distribution. Additionally, we integrate the global representation of the feature map obtained from the Transformer with the locally extracted information via convolutional layers. This integration achieves a harmonious synergy that optimizes the diagnostic prowess of the model. Main results. Overall, the accuracy, specificity, F1-Score, and precision in the experimental results were 98.38%, 99.04%, 98.66% and 99.38%, respectively. Significance. This study introduces a deep learning network equipped with both global feature response and local feature extraction capabilities. This innovative architecture holds significant promise for the accurate diagnosis of CD by leveraging endoscopic images captured from diverse anatomical sites.

A hierarchical federated learning framework for collaborative quality defect inspection in construction

Recent advancements in robotics and deep learning (DL) have made it possible to implement robots in civil infrastructures' quality defect inspection. Robots can reduce human inspectors' workloads and enhance inspection results' reliability by collecting data and automatically identifying quality defects from the raw data. However, current methods for training DL models rely on centralized strategies that require the aggregation of defect data (e.g., uploading to a cloud server), posing concerns about data privacy and security. Thus, this study proposes a three-fold federated learning (FL) framework for training DL models collaboratively, without the need to share local data among construction robots. The framework is specifically applied to image-based crack segmentation, critical for ensuring infrastructures’ safety and serviceability. A lightweight DL model is developed to enable easy implementation on resource-constrained construction robots and to reduce communication costs during federated training. Experimental results show that the proposed FL method outperforms traditional centralized methods. The critical contribution of this study is the hierarchical FL framework, which enables construction robots to leverage big data in a privacy-preserving manner.

MMDB: Multimodal dual-branch model for multi-functional bioactive peptide prediction

Bioactive peptides can hinder oxidative processes and microbial spoilage in foodstuffs and play important roles in treating diverse diseases and disorders. While most of the methods focus on single-functional bioactive peptides and have obtained promising prediction performance, it is still a significant challenge to accurately detect complex and diverse functions simultaneously with the quick increase of multi-functional bioactive peptides. In contrast to previous research on multi-functional bioactive peptide prediction based solely on sequence, we propose a novel multimodal dual-branch (MMDB) lightweight deep learning model that designs two different branches to effectively capture the complementary information of peptide sequence and structural properties. Specifically, a multi-scale dilated convolution with Bi-LSTM branch is presented to effectively model the different scales sequence properties of peptides while a multi-layer convolution branch is proposed to capture structural information. To the best of our knowledge, this is the first effective extraction of peptide sequence features using multi-scale dilated convolution without parameter increase. Multimodal features from both branches are integrated via a fully connected layer for multi-label classification. Compared to state-of-the-art methods, our MMDB model exhibits competitive results across metrics, with a 9.1% Coverage increase and 5.3% and 3.5% improvements in Precision and Accuracy, respectively.

A Lightweight Deep Learning Model and Web Interface for COVID-19 Detection Using Chest X-Rays

A Lightweight Deep Learning Model and Web Interface for COVID-19 Detection Using Chest X-Rays

Pattern Augmented Lightweight Convolutional Neural Network for Intrusion Detection System

As the world increasingly becomes more interconnected, the demand for safety and security is ever-increasing, particularly for industrial networks. This has prompted numerous researchers to investigate different methodologies and techniques suitable for intrusion detection systems (IDS) requirements. Over the years, many studies have proposed various solutions in this regard, including signature-based and machine learning (ML)-based systems. More recently, researchers are considering deep learning (DL)-based anomaly detection approaches. Most proposed works in this research field aim to achieve either one or a combination of high accuracy, considerably low false alarm rates (FARs), high classification specificity and detection sensitivity, lightweight DL models, or other ML and DL-related performance measurement metrics. In this study, we propose a novel method to convert a raw dataset to an image dataset to magnify patterns by utilizing the Short-Term Fourier transform (STFT). The resulting high-quality image dataset allowed us to devise an anomaly detection system for IDS using a simple lightweight convolutional neural network (CNN) that classifies denial of service and distributed denial of service. The proposed methods were evaluated using a modern dataset, CSE-CIC-IDS2018, and a legacy dataset, NSLKDD. We have also applied a combined dataset to assess the generalization of the proposed model across various datasets. Our experimental results have demonstrated that the proposed methods achieved high accuracy and considerably low FARs with high specificity and sensitivity. The resulting loss and accuracy curves have demonstrated the efficacy of our raw dataset to image dataset conversion methodology, which is evident as an excellent generalization of the proposed lightweight CNN model was observed, effectively avoiding overfitting. This holds for both the modern and legacy datasets, including their mixed versions.

Feasibility of video-based real-time nystagmus tracking: a lightweight deep learning model approach using ocular object segmentation.

Eye movement tests remain significantly underutilized in emergency departments and primary healthcare units, despite their superior diagnostic sensitivity compared to neuroimaging modalities for the differential diagnosis of acute vertigo. This underutilization may be attributed to a potential lack of awareness regarding these tests and the absence of appropriate tools for detecting nystagmus. This study aimed to develop a nystagmus measurement algorithm using a lightweight deep-learning model that recognizes the ocular regions. The deep learning model was used to segment the eye regions, detect blinking, and determine the pupil center. The model was trained using images extracted from video clips of a clinical battery of eye movement tests and synthesized images reproducing real eye movement scenarios using virtual reality. Each eye image was annotated with segmentation masks of the sclera, iris, and pupil, with gaze vectors of the pupil center for eye tracking. We conducted a comprehensive evaluation of model performance and its execution speeds in comparison to various alternative models using metrics that are suitable for the tasks. The mean Intersection over Union values of the segmentation model ranged from 0.90 to 0.97 for different classes (sclera, iris, and pupil) across types of images (synthetic vs. real-world images). Additionally, the mean absolute error for eye tracking was 0.595 for real-world data and the F1 score for blink detection was ≥ 0.95, which indicates our model is performing at a very high level of accuracy. Execution speed was also the most rapid for ocular object segmentation under the same hardware condition as compared to alternative models. The prediction for horizontal and vertical nystagmus in real eye movement video revealed high accuracy with a strong correlation between the observed and predicted values (r = 0.9949 for horizontal and r = 0.9950 for vertical; both p < 0.05). The potential of our model, which can automatically segment ocular regions and track nystagmus in real time from eye movement videos, holds significant promise for emergency settings or remote intervention within the field of neurotology.

IPMCNet: A Lightweight Algorithm for Invasive Plant Multiclassification

Invasive plant species pose significant biodiversity and ecosystem threats. Real-time identification of invasive plants is a crucial prerequisite for early and timely prevention. While deep learning has shown promising results in plant recognition, the use of deep learning models often involve a large number of parameters and high data requirements for training. Unfortunately, the available data for various invasive plant species are often limited. To address this challenge, this study proposes a lightweight deep learning model called IPMCNet for the identification of multiple invasive plant species. IPMCNet attains high recognition accuracy even with limited data and exhibits strong generalizability. Simultaneously, by employing depth-wise separable convolutional kernels, splitting channels, and eliminating fully connected layer, the model’s parameter count is lower than that of some existing lightweight models. Additionally, the study explores the impact of different loss functions, and the insertion of various attention modules on the model’s accuracy. The experimental results reveal that, compared with eight other existing neural network models, IPMCNet achieves the highest classification accuracy of 94.52%. Furthermore, the findings suggest that focal loss is the most effective loss function. The performance of the six attention modules is suboptimal, and their insertion leads to a decrease in model accuracy.

Light-Weight Deep Learning Model for Accelerating the Classification of Mango-Leaf Disease

Mango leaf diseases represent a serious threat to world agriculture, necessitating prompt and accurate detection to avert catastrophic effects. In response, this study suggests a light-weight, deep learning-based method for automatically classifying mango leaf diseases. The model is based on the original DenseNet architecture, which is well known for its effectiveness in image classification tasks. Custom layers have been added over the existing layer of the original DenseNet model. The proposed model has been compared with other existing pre-trained models. Based on comparisons, the proposed model, DenseNet78, proved to be efficient even on a relatively small dataset, where the conventional model failed. The proposed model ensured generalization across regions, disease variants, and diverse datasets of mango leaves. The results demonstrate that the fine-tuned DenseNet architecture (DenseNet78), along with an ideal growth rate, modifying block size, and a number of layers, provides optimum accuracy, with 99.47% accuracy in identifying healthy mango leaves and 99.44% accuracy in detecting various mango leaf diseases. The results also demonstrate that the model is effective in accelerating the training process because of careful comparative analysis of all the available alternatives, including the most effective combination of optimizers, learning rate schedulers, and loss functions. The study's conclusion is an automated approach for diagnosing mango leaf disease using an improved and optimized DenseNet architecture (DenseNet78). Doi: 10.28991/ESJ-2024-08-01-03 Full Text: PDF

An IoT Transfer Learning-Based Service for the Health Status Monitoring of Grapevines

Grapevine is a valuable and profitable crop that is susceptible to various diseases, making effective disease detection crucial for crop monitoring. This work explores the use of deep learning-based plant disease detection as an alternative to traditional methods, employing an Internet of Things approach. An edge device, a Raspberry Pi 4 equipped with an RGB camera, is utilized to detect diseases in grapevine plants. Two lightweight deep learning models, MobileNet V2 and EfficientNet B0, were trained using a transfer learning technique on commercially available online dataset, then deployed and validated on field-site in an organic winery. The models’ performance was further enhanced using semantic segmentation with the Mobile-UNet algorithm. Results were reported through a web service using FastAPI. Both models achieved high training accuracies exceeding 95%, with MobileNet V2 slightly outperforming EfficientNet B0. During validation, MobileNet V2 achieved an accuracy of 94%, compared to 92% for EfficientNet B0. In terms of IoT deployment, MobileNet V2 exhibits faster inference time (330 ms) compared to EfficientNet B0 (390 ms), making it the preferred model for online deployment.

TinyML: Adopting tiny machine learning in smart cities

&lt;div&gt;&lt;p&gt;Since Tiny machine learning (TinyML) is a quickly evolving subject, it is crucial that internet of things (IoT) devices be able to communicate with one another for the sake of stability and future development. TinyML is a rapidly growing subfield at the intersection of computer science, software engineering, and machine learning. Building deep learning (DL) networks that are a few hundred KBs in size has been the focus of recent research in this area. Deploying TinyML into small devices makes them smart. Reduced computation, power usage, and response time are just a few of the many advantages of TinyML. In this work, we provide an introduction to TinyML and demonstrate its benefits and architecture. Then, we investigate the meaning of quantization as a standard compression method for TinyML-related applications. There are two methods used to obtain the quantized weights of the deep learning models are quantization-aware training (QAT) and post-training quantization (PTQ), we described them in details. Next, TinyML-based solutions to improve the role of IoT devices in Smart Cities are highlighted as: lightweight training of deep learning models, inference of lightweight deep learning models in IoT devices, low power consumption, and inference of deep learning models in restricted resources of IoT devices. Finally, presenting some use cases for TinyML studies, with these studies applied to several cases in a variety of fields. To the best of the author’s knowledge, few studies have investigated TinyML as it is an emerging field.&lt;/p&gt;&lt;/div&gt;

Unsupervised Deep Anomaly Detection for Industrial Multivariate Time Series Data

With the rapid development of deep learning, researchers are actively exploring its applications in the field of industrial anomaly detection. Deep learning methods differ significantly from traditional mathematical modeling approaches, eliminating the need for intricate mathematical derivations and offering greater flexibility. Deep learning technologies have demonstrated outstanding performance in anomaly detection problems and gained widespread recognition. However, when dealing with multivariate data anomaly detection problems, deep learning faces challenges such as large-scale data annotation and handling relationships between complex data variables. To address these challenges, this study proposes an innovative and lightweight deep learning model—the Attention-Based Deep Convolutional Autoencoding Prediction Network (AT-DCAEP). The model consists of a characterization network based on convolutional autoencoders and a prediction network based on attention mechanisms. The AT-DCAEP exhibits excellent performance in multivariate time series data anomaly detection without the need for pre-labeling large-scale datasets, making it an efficient unsupervised anomaly detection method. We extensively tested the performance of AT-DCAEP on six publicly available datasets, and the results show that compared to current state-of-the-art methods, AT-DCAEP demonstrates superior performance, achieving the optimal balance between anomaly detection performance and computational cost.