Lightweight Network Model Research Articles

MPE-HRNetL: A Lightweight High-Resolution Network for Multispecies Animal Pose Estimation

Animal pose estimation is crucial for animal health assessment, species protection, and behavior analysis. It is an inevitable and unstoppable trend to apply deep learning to animal pose estimation. In many practical application scenarios, pose estimation models must be deployed on edge devices with limited resource. Therefore, it is essential to strike a balance between model complexity and accuracy. To address this issue, we propose a lightweight network model, i.e., MPE-HRNet.L, by improving Lite-HRNet. The improvements are threefold. Firstly, we improve Spatial Pyramid Pooling-Fast and apply it and the improved version to different branches. Secondly, we construct a feature extraction module based on a mixed pooling module and a dual spatial and channel attention mechanism, and take the feature extraction module as the basic module of MPE-HRNet.L. Thirdly, we introduce a feature enhancement stage to enhance important features. The experimental results on the AP-10K dataset and the Animal Pose dataset verify the effectiveness and efficiency of MPE-HRNet.L.

LRSDet: lightweight remote sensing target detection network with local-global information awareness and rapid sample assignment

Abstract With the rapid development of aerospace and unmanned aerial vehicles, using neural networks for object detection in optical remote sensing images (O-RSI) has encountered heightened challenges. The optical remote sensing images have the characteristics of complex geometric scenes, dense groups of objects, and significant multi-scale variations of objects; researchers need to use more complex models to achieve higher accuracy. However, this complexity also brings challenges to the application of lightweight scenes. Therefore, to cope with the trade-off challenge between model complexity and detection accuracy, we propose a lightweight network model LRSDet in this study. The model integrates local and global information processing mechanisms and introduces a fast positive sample assignment strategy to adapt to resource-constrained embedded and mobile platforms. By constructing a lightweight feature extraction network and a lightweight path aggregation network and incorporating the ESM-Attention module, the feature extraction capability of the model in complex remote sensing scenarios is significantly improved. In addition, the application of a dynamic soft threshold strategy further optimizes the positive sample selection process and improves the detection efficiency of the model. Experimental on the O-RSI datasets DIOR, NWPU VHR-10, and RSOD, while analyzing model real-time performance on aerial video and embedded devices, outperforming other state-of-the-art methods.

Lightweight UAV target detection algorithm

Abstract Aiming at the public area UAV detection task, the target is easy to occlude, the scale variation is large, and the existing detection algorithm model parameters are large. In this paper, the lightweight network model YOLOv5s-UD is designed for the low-altitude UAV target detection algorithm. In this detection model, the proposed approach employs the streamlined Shufflenetv2 architecture to enhance the core network. To optimize the UAV’s relevant feature data, the CBAM attention module is integrated during the feature amalgamation phase. Furthermore, the GSConv is implemented to upgrade conventional convolution within the neck network, ensuring a reduction in model parameters without compromising detection precision. Experimental outcomes reveal a 36.2% computational reduction compared to the YOLOv5s algorithm, with an average detection accuracy of 84.1%. This demonstrates the algorithm’s capability to efficiently detect UAV targets while minimizing computational loads, thereby exhibiting significant practical significance.

MobilenetV2-RC: A lightweight network model for retinopathy classification in retinal OCT images

Abstract Retinopathy is an important ophthalmic disease that causes blindness in the elderly population. With the global old peoples increasing, the ophthalmic healthcare system is more and significant to pre-diagnosis. Optical Coherence Tomography (OCT) is considered the gold standard for ophthalmic treatment and diagnosis. OCT technologies and equipment continue to develop towards the intelligence and convenience for requirements of rapid diagnosis in the remote and poverty-stricken areas. We proposed an improved MobilenetV2 lightweight model for retinopathy classification (MobilenetV2-RC), which incorporates Spatial and Channel Reconstruction Convolution (SCConv) and the improved Convolutional Block Attention Module (CBAM) attention mechanism into the framework. Not only can it effectively limit feature redundancy to reduce model parameters, but also enhance the ability of feature representation to improve classification accuracy. The parameters of the proposed model are only 1.96 M with an overall accuracy of 98.96%, which is higher 3.32% than the original MobilenetV2. Compared with ResNet18, InceptionV3, and VGG16_BN, the overall accuracy is increased by 4.6%, 6.3%, and 3.9%, respectively. The test results of UCSD and Duke open-source datasets are more remarkable. Experimental results show that our proposed algorithm has strong reliability and generalization for the accurate classification of retinopathy, and a greater application prospect in the intelligent diagnosis of ophthalmology and mobile detection terminals.

A multi-channel spatial information feature based human pose estimation algorithm

Human pose estimation is an important task in computer vision, which can provide key point detection of human body and obtain bone information. At present, human pose estimation is mainly utilized for detection of large targets, and there is no solution for detection of small targets. This paper proposes a multi-channel spatial information feature based human pose (MCSF-Pose) estimation algorithm to address the issue of medium and small targets inaccurate detection of human key points in scenarios involving occlusion and multiple poses. The MCSF-Pose network is a bottom-up regression network. Firstly, an UP-Focus module is designed to expand the feature information while reducing parameter computation during the up-sampling process. Then, the channel segmentation strategy is adopted to cut the features, and the feature information of multiple dimensions is retained through different convolutional groups, which reduces the parameter lightweight network model and makes up for the loss of the feature information associated with the depth of the network. Finally, the three-layer PANet structure is designed to reduce the complexity of the model. With the aid of the structure, it also to improve the detection accuracy and anti-interference ability of human key points. The experimental results indicate that the proposed algorithm outperforms YOLO-Pose and other human pose estimation algorithms on COCO2017 and MPII human pose datasets.

A method of dense point cloud SLAM based on improved YOLOV8 and fused with ORB-SLAM3 to cope with dynamic environments

With the development of society and the advancement of technology, intelligent robots have been widely used in various fields. At the same time, Simultaneous Localization and Mapping (SLAM) technology is a key technology in the research field of intelligent robots. However, in dynamic environments, achieving accurate and robust visual SLAM remains a major challenge. In this paper, we propose a method based on improved YOLOv8 fused with ORB-SLAM3 to address dense point cloud SLAM in dynamic environments. Our proposed method successfully integrates real-time object detection and image segmentation technologies of YOLOv8 into the ORB-SLAM3 framework, achieving high-precision and robust visual SLAM in dynamic environments. In the YOLOv8 framework, we use a balanced convolution method, GSConv, instead of some traditional convolution layers (Conv), which balances accuracy with computational load. Based on the GSConv convolution method, we adopt a new feature fusion module, VoVGSCSP, to replace traditional C2f feature fusion modules, thereby improving the Neck structure of YOLOv8 and achieving a lightweight network model. We compare our proposed method with ORB-SLAM3 and some computer vision algorithms on the TUM dataset. Experimental data confirms that our method outperforms existing visual SLAM algorithms in dynamic environments. In fast-moving dynamic environments, the RMSE of absolute pose estimation of our method is 96.28% lower than that of ORB-SLAM3, and the RMSE of relative pose estimation is 51.57% lower than that of ORB-SLAM3. The experimental results demonstrate that our method significantly improves the accuracy of pose estimation in dynamic environments and greatly enhances the performance compared to ORB-SLAM3.

Data Processing of Bridge IoT Monitoring Based on Task Scheduling of Cloud Service Listening Signal

For the long-term continuous monitoring of bridge-related indicators, it is necessary to arrange relatively perfect acquisition equipment on the bridge, which can feedback various information parameters of the bridge. However, there are many parameters to feedback the bridge information, which leads to the complex and overstaffed structure of the monitoring system. Furthermore, the huge amount of data collected and the complex calculation process also increase the difficulty of the operation of the monitoring system. In this regard, we should choose more scientific and reasonable indicators, lightweight data structure, stable data transmission, and analysis programs to improve the accuracy of continuous monitoring. To establish a stable and efficient bridge monitoring system, we use the distance coefficient-effective independent algorithm to optimize. Then, we calculate the relevant information of the strain environment with the help of a neural network model, strengthen the training of deep learning through the YOLOv5s model, and improve the task scheduling strategy of attention concentration. Through that, we solve the problem of embedded systems with relatively low computing power. Different weights are assigned to each fused feature map, and the nodes at the highest level and the lowest level are deleted so that a concise and efficient lightweight network model is constructed. Multiple iterations are performed to achieve deeper feature fusion. Therefore, the complexity of the model is effectively reduced, and the monitoring performance can be effectively improved. Finally, through the experimental analysis, it is proved that compared with the traditional fusion model, the number of parameters of the improved fusion network structure in bridge health monitoring is reduced by 7.37%. The detection speed is increased by 18.2%. The amount of computation is reduced by 42.92%, and the average detection accuracy is required to reach 95.33%. It is verified that the proposed method can effectively improve the accuracy and risk control ability of the detection data by learning from the samples with small labels. It also has great practical significance and market value for the design and optimization of the bridge health monitoring system, which is suitable for the monitoring data of large-scale construction projects.

Lightweight-detection: The strip steel surface defect identification based on improved YOLOv5d

The surface defect of strip steel seriously affects the product quality, the current identification methods have the inherent defects of low detection efficiency and poor detection accuracy, and it is important an improved an algorithm based on the modified YOLOv5 framework to enhance the defect detection effectiveness and accuracy. To create a more lightweight network model, the C3 module and part of the convolutional structure in the original YOLOv5 network were replaced with the GhostBottleneck structure. Additionally, the Coordinate Attention (CA) mechanism was incorporated into the Backbone section to compensate for the original model's lack of attention mechanisms. To address the angular mismatch between the actual frame and the predicted bounding box, a limitation overlooked by the CIOU loss function was further refined. Experimental results demonstrate that the enhanced YOLOv5s-CSG model outperformed the original YOLOv5s by 7.3 %, achieving a significant 37.9 % reduction in model size and an mAP value of 77.5 %. Furthermore, the detection precision and speed were notably higher compared to several widely-used algorithms. The proposed YOLOv5s-CSG model is well-suited for deployment on mobile devices and shows great potential for rapidly and accurately detecting both the type and location of strip surface defects, surpassing the performance of existing methods.

EL-Net: An efficient and lightweight optimized network for object detection in remote sensing images

Object detection in Unmanned Aerial Vehicles (UAV) optical remote sensing imagery presents a formidable challenge in computer vision due to the minuscule size of targets, which occupy fewer pixels and provide limited feature information, complicating accurate recognition and classification. Furthermore, the overlapping of dense targets exacerbates the difficulty of precise classification and localization. Meanwhile, classical detection networks often struggle to balance recognition accuracy with model complexity. Addressing these issues, this paper introduces EL-Net, an efficient and lightweight network model based on improvements to the YOLOv7-tiny architecture. First, the network structure is streamlined through a lightweight design that maintains performance while reducing complexity. Additionally, a feature perception enhancement module (FPEM) using attention mechanisms and dilated convolution significantly improves the model’s capability to extract key features from complex backgrounds. Finally, the optimized network structure is compressed by a structured pruning algorithm. EL-Net was evaluated in challenging scenarios on the VisDrone2019 dataset, where it achieved a mean Average Precision (mAP) of 38.7%, demonstrating high detection accuracy at minimal model complexity. Meanwhile, evaluation of the UA-DETRAC dataset has demonstrated the model’s remarkable generalization capacity. The outcomes suggest that EL-Net effectively balances accuracy and efficiency, making it ideal for deployment on resource-limited mobile edge devices while offering an innovative approach to object detection in UAV imagery.

YOLO-Chili: An Efficient Lightweight Network Model for Localization of Pepper Picking in Complex Environments

Currently, few deep models are applied to pepper-picking detection, and existing generalized neural networks face issues such as large model parameters, prolonged training times, and low accuracy. To address these challenges, this paper proposes the YOLO-chili target detection algorithm for chili pepper detection. Initially, the classical target detection algorithm YOLOv5 serves as the benchmark model. We introduce an adaptive spatial feature pyramid structure that combines the attention mechanism and the concept of multi-scale prediction to enhance the model’s detection capabilities for occluded and small target peppers. Subsequently, we incorporate a three-channel attention mechanism module to improve the algorithm’s long-distance recognition ability and reduce interference from redundant objects. Finally, we employ a quantized pruning method to reduce model parameters and achieve lightweight processing. Applying this method to our custom chili pepper dataset, we achieve an average precision (AP) value of 93.11% for chili pepper detection, with an accuracy rate of 93.51% and a recall rate of 92.55%. The experimental results demonstrate that YOLO-chili enables accurate and real-time pepper detection in complex orchard environments.

A Forest Fire Smoke Monitoring System Based on a Lightweight Neural Network for Edge Devices

Forest resources are one of the indispensable resources of the earth, which are the basis for the survival and development of human society. With the swift advancements in computer vision and artificial intelligence technology, the utilization of deep learning for smoke detection has achieved remarkable results. However, the existing deep learning models have poor performance in forest scenes and are difficult to deploy because of numerous parameters. Hence, we introduce an optimized forest fire smoke monitoring system for embedded edge devices based on a lightweight deep learning model. The model makes full use of the multi-scale variable attention mechanism of Transformer architecture to strengthen the ability of image feature extraction. Considering the needs of application scenarios, we propose an improved lightweight network model LCNet for feature extraction, which can reduce the parameters and enhance detecting ability. In order to improve running speed, a simple semi-supervised label knowledge distillation scheme is used to enhance the overall detection capability. Finally, we design and implement a forest fire smoke detection system on an embedded device, including the Jetson NX hardware platform, high-definition camera, and detection software system. The lightweight model is transplanted to the embedded edge device to achieve rapid forest fire smoke detection. Also, an asynchronous processing framework is designed to make the system highly available and robust. The improved model reduces three-fourths of the parameters and increases speed by 3.4 times with similar accuracy to the original model. This demonstrates that our system meets the precision demand and detects smoke in time.

A high-accuracy lightweight network model for X-ray image diagnosis: A case study of COVID detection.

The Coronavirus Disease 2019(COVID-19) has caused widespread and significant harm globally. In order to address the urgent demand for a rapid and reliable diagnostic approach to mitigate transmission, the application of deep learning stands as a viable solution. The impracticality of many existing models is attributed to excessively large parameters, significantly limiting their utility. Additionally, the classification accuracy of the model with few parameters falls short of desirable levels. Motivated by this observation, the present study employs the lightweight network MobileNetV3 as the underlying architecture. This paper incorporates the dense block to capture intricate spatial information in images, as well as the transition layer designed to reduce the size and channel number of the feature map. Furthermore, this paper employs label smoothing loss to address the inter-class similarity effects and uses class weighting to tackle the problem of data imbalance. Additionally, this study applies the pruning technique to eliminate unnecessary structures and further reduce the number of parameters. As a result, this improved model achieves an impressive 98.71% accuracy on an openly accessible database, while utilizing only 5.94 million parameters. Compared to the previous method, this maximum improvement reaches 5.41%. Moreover, this research successfully reduces the parameter count by up to 24 times, showcasing the efficacy of our approach. This demonstrates the significant benefits in regions with limited availability of medical resources.

Research on lightweight speech emotion recognition in vehicle noisy environment

In order to reduce the incidence of traffic accidents caused by the emotional state of drivers, this study proposes an emotion recognition algorithm based on vehicle noise environment. This algorithm can effectively identify the emotional state of drivers and provide support for further improving their emotions. To address challenges in existing research on speech emotion recognition, such as excessive model parameters, poor generalization, and suboptimal performance in noisy environments, this paper proposes a lightweight network model suitable for small datasets. The model utilizes Power Normalized Cepstral Coefficients (PNCC) as input features, and employs parallel feature extraction layers at different scales. These features are then fed into a feature learning module for in-depth extraction, with the final determination of the driver’s emotional state made by the output layer. Experimental results show that the model achieves an accuracy of 96.08% on the EMO-DB speech dataset. Even in simulated in-vehicle noise environments, the model exhibits high accuracy and robustness. Moreover, compared to other lightweight models, it has fewer training parameters and faster processing speed, making it suitable for deployment on edge devices in mobile applications.

A lightweight network model designed for alligator gar detection

When using advanced detection algorithms to monitor alligator gar in real-time in wild waters, the efficiency of existing detection algorithms is subject to certain limitations due to turbid water quality, poor underwater lighting conditions, and obstruction by other objects. In order to solve this problem, we developed a lightweight real-time detection network model called ARD-Net, from the perspective of reducing the amount of calculation and obtaining more feature map patterns. We introduced a cross-domain grid matching strategy to accelerate network convergence, and combined the involution operator and dual-channel attention mechanism to build a more lightweight feature extractor and multi-scale detection reasoning network module to enhance the network’s response to different semantics. Compared with the yoloV5 baseline model, our method performs equivalently in terms of detection accuracy, but the model is smaller, the detection speed is increased by 1.48 times, When compared with the latest State-of-the-Art (SOTA) method, YOLOv8, our method demonstrates clear advantages in both detection efficiency and model size,and has good real-time performance. Additionally, we created a dataset of alligator gar images for training.

Autonomous surface crack identification for concrete structures based on the you only look once version 5 algorithm

Failure to repair roads in a timely manner may shorten their life and even cause traffic accidents. Thus, accurate crack detection and reasonable classification are crucial for road safety evaluation. In this study, an improved network model based on the You Only Look Once version 5 algorithm is presented, with three additional modules: The first module improves the data processing speed by replacing the C3 module in the original network with a lightweight network model. The second module was used to lighten network weight by reusing a simple convolution structure to equivalently represent the calculation of a convolution layer as a weighted sum of several small convolution blocks. And the third module is used to improve the detection accuracy by removing the upsampling and performing three-way splicing. The proposed model can detect different types of cracks, and an extensive ablation study is reported based on various combinations of the proposed modules. Based on training on a database of 5484 images, the results show that the improved network proposed in this study can effectively identify pavement cracks. Compared with the original network, mean Average Precision is increased by 5.98%, the inference time is reduced by 4.82%, and the model weight is decreased by 17.36%. Additionally, to comply with engineering practice, comparative experiments were conducted on the pre-rotated dataset. The results showed that compared with You Only Look Once version 8, the improved algorithm improved accuracy, recall, average accuracy, and F1 score by 3.28%, 8.46%, 3.79% and 5.89%, respectively. This study can serve as an important reference for the development of crack detection methods.

Defect identification of electricity transmission line insulators based on the improved lightweight network model with computer vision assistance

This work aims to ensure the safe operation of electricity transmission lines and reduce costs and maintenance difficulties. It studies the application of computer vision (CV) in the defect identification of electricity transmission lines. In addition, this work proposes a method to improve the lightweight network model to provide an effective identification model to solve the problem of electricity transmission line defects. Firstly, GraphCut segmentation and Laplace algorithms are employed to expand and sharpen the electricity transmission line image. Secondly, in light of the Depth Separable Convolution algorithm, a defect detection model for the electricity transmission line insulator is proposed based on the You Only Look Once 4 (YOLOv4) network. Moreover, MobileNetV1 is utilized to improve this lightweight network model. Finally, this work uses ImageNet, a large public dataset, to validate the proposed model experimentally. The research results reveal that: (1) In the model testing results, all research indicators of the model are greater than 90 %, indicating an excellent detection accuracy of this model. (2) The improved YOLOv4 model can increase the detection speed to 53 frames/s at the cost of 2.4 % accuracy. (3) After image sharpening, the improved YOLOv4 model has promoted the insulator defects' detection ability to a certain extent. The above outcomes suggest that the improved YOLOv4 model can predict more efficiently and accurately and reduce unnecessary false positives. This illustrates that the proposed model is feasible and is expected to be applied to the defect identification of electricity transmission lines in practice. These findings fully demonstrate this work's vital value in enhancing the prediction efficiency and accuracy, thus offering a strong preference for the defect identification of electricity transmission lines in practical applications.

Effective Cataract Identification System using Deep Convolution Neural Network

INTRODUCTION: The paper introduces a novel approach for the early detection of cataracts using images captured using smartphones. Cataracts are a significant global eye disease that can lead to vision impairment in individuals aged 40 and above. In this article, we proposed a deep convolution neural network (CataractsNET) trained using an open dataset available in Github which includes images collected through google searches and images generated using standard augmentation mechanism.
 OBJECTIVES: The main objective of this paper is to design and implement a lightweight network model for cataract identification that outperforms other state-of-the-art network models in terms of accuracy, precision, recall, and F1 Score.
 METHODS: The proposed neural network model comprises nine layers, guaranteeing the extraction of significant details from the input images and achieving precise classification. The dataset primarily comprises cataract images sourced from a standardized dataset that is publicly available on GitHub, with 8000 training images and 1600 testing images.
 RESULTS: The proposed CataractsNET model achieved an accuracy of 96.20%, precision of 96.1%, recall of 97.6%, and F1 score of 96.1%. These results demonstrate that the proposed method outperforms other deep learning models like ResNet50 and VGG19.
 CONCLUSION: The paper concludes that identifying cataracts in the earlier stages is crucial for effective treatment and reducing the likelihood of experiencing blindness. The widespread use of smartphones makes this approach accessible to a broad audience, allowing individuals to check for cataracts and seek timely consultation with ophthalmologists for further diagnosis.

Driver Abnormal Expression Detection Method Based on Improved Lightweight YOLOv5

The rapid advancement of intelligent assisted driving technology has significantly enhanced transportation convenience in society and contributed to the mitigation of traffic safety hazards. Addressing the potential for drivers to experience abnormal physical conditions during the driving process, an enhanced lightweight network model based on YOLOv5 for detecting abnormal facial expressions of drivers is proposed in this paper. Initially, the lightweighting of the YOLOv5 backbone network is achieved by integrating the FasterNet Block, a lightweight module from the FasterNet network, with the C3 module in the main network. This combination forms the C3-faster module. Subsequently, the original convolutional modules in the YOLOv5 model are replaced with the improved GSConvns module to reduce computational load. Building upon the GSConvns module, the VoV-GSCSP module is constructed to ensure the lightweighting of the neck network while maintaining detection accuracy. Finally, channel pruning and fine-tuning operations are applied to the entire model. Channel pruning involves removing channels with minimal impact on output results, further reducing the model’s computational load, parameters, and size. The fine-tuning operation compensates for any potential loss in detection accuracy. Experimental results demonstrate that the proposed model achieves a substantial reduction in both parameter count and computational load while maintaining a high detection accuracy of 84.5%. The improved model has a compact size of only 4.6 MB, making it more conducive to the efficient operation of onboard computers.

Usage of an improved YOLOv5 for steel surface defect detection

Abstract The one-stage YOLOv5 steel surface defect detection has issues such as slow operation speed, loss of defect location and semantic information of small targets, and inadequate extraction of defect features. This study proposed a defect detection algorithm with improved YOLOv5 to solve these issues. The proposed algorithm used the slim-neck layer built by three new modules instead of the neck layer in YOLOv5s to achieve a lightweight network model. In addition, the spatial perception self-attention mechanism was introduced to enhance the feature extraction capability of the initial convolutional layer without limiting the input size. The improved Atrous Spatial Pyramid Pooling was added to expand the perceptual field and capture multiscale contextual information while preventing local information loss and enhancing the relevance of long-range information. The experimental results showed that the improved YOLOv5 algorithm has a reduced model volume, significantly higher detection accuracy and speed than the traditional algorithm, and the ability to detect steel surface defects quickly and accurately.

Wood identification based on macroscopic images using deep and transfer learning approaches.

Identifying forest types is vital for evaluating the ecological, economic, and social benefits provided by forests, and for protecting, managing, and sustaining them. Although traditionally based on expert observation, recent developments have increased the use of technologies such as artificial intelligence (AI). The use of advanced methods such as deep learning will make forest species recognition faster and easier. In this study, the deep network models RestNet18, GoogLeNet, VGG19, Inceptionv3, MobileNetv2, DenseNet201, InceptionResNetv2, EfficientNet and ShuffleNet, which were pre-trained with ImageNet dataset, were adapted to a new dataset. In this adaptation, transfer learning method is used. These models have different architectures that allow a wide range of performance evaluation. The performance of the model was evaluated by accuracy, recall, precision, F1-score, specificity and Matthews correlation coefficient. ShuffleNet was proposed as a lightweight network model that achieves high performance with low computational power and resource requirements. This model was an efficient model with an accuracy close to other models with customisation. This study reveals that deep network models are an effective tool in the field of forest species recognition. This study makes an important contribution to the conservation and management of forests.