Small target detection algorithm based on YOLOv4

Li, Quanfang; Zhu, Liucun; Li, Boquan; Luo, Junqi; Liu, Daopeng; Chen, Mingyou

doi:10.1109/scset55041.2022.00026

Cited by 6 publications

(9 citation statements)

References 2 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The proposed method investigates the clusters with the performance metrics given in the equations 5 to 7, Comparing with the existing methods which are discussed in the review literature. (2) 3 75.34 84.24 83.56 DBSCAN (5) 5,7 83.15 89.00 90.31 GSVM (6) 5 79.76 90.45 92.22 SOM (8) 5 75.00 78.73 93.51 PCA (11) 3, 5 85.69 94.88 93.11 ocSVM (15) 3, 5 80.23 92.57 90.55 ICA (20) 3 From Table 3, K-Means process the three different cluster classes to find the accuracy level to 83.56%, the energy consumption rate is low to compute the cluster set. But the Algorithm improves the classification to 84.24% comparing the other network functionality to fabricate the attacking nodes.…”

Section: Resultsmentioning

confidence: 99%

“…The network analytics process to find the decaying network entities avoiding the normal flow and progress of network utilized for a variety of purposes, including finding connected endpoints, identifying bottlenecks, assessing device health, resolving issues, and probing for potential security flaws also determines how to optimize operations by comparing incoming data with pre-programmed models (1) . The aggregation approaches often makes discrete steps with time interval steps with a single piece of linear function or mixture of functions significantly on samples per object than a greater number of objects (2) . To finds the lag of service time the network functionalities directly measure similarity between the synchronizing methods or the methods that aggregate multiple imputations used for clustering time-series of data with the transmission constraints (3) .…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Intellectual Data Aggregation Using Independent Cluster based Medicaid Method for Network Functionality Fabrication

Moorthy¹,

Shanthakumar²,

Janarthanam³

2022

IJST

View full text Add to dashboard Cite

Objectives: Network functionality-based tracking is difficult for the fabrication processing under the various environment condition. In order to achieve the efficient result Intellectual data cluster-based Medicaid aggregation method applied for the network functionality analysis of the attack on network entities aggregate the fabrication process for balancing the network to normal flow. Methods: The proposed Independent Cluster based Medicaid Intellectual Data Aggregation (ICAMA) is compared with the K-means and GSVM, DBSCAN algorithm for classification of the attack entities to understand the response and reinforcement of the network functionality against the interconnection nodes by using energy and weighted aggregation, accuracy metrics on the CIC-IDS 2019 dataset. Findings: The Experiment studies provide the 15.23% lower energy consumption for interconnection of nodes comparing with the DBSCAN and 17.18% against the K-means algorithms. The network functional fabrication progress of proposed method has the weighted aggregation15.07%,8.32%, 6.41%, 5.12%, 5.52%, 8.08% and 2.21% against the given K-means, DBSCAN, GSVM, SOM, PCA, ocSVM and ICA methods and also improves accuracy to 11.07% compared with the GSVM method. ICAMA applies the intellectual functionality based on the distribution of entities in the network as a cluster or a single fabrication process. Novelty: In order to restore the regular network progress, ICAMA continuously runs the dynamic collection of network analysis by employing intellectual functional support to identify various types of network anomalies.

show abstract

Section: Resultsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Intellectual Data Aggregation Using Independent Cluster based Medicaid Method for Network Functionality Fabrication

Moorthy¹,

Shanthakumar²,

Janarthanam³

2022

IJST

View full text Add to dashboard Cite

show abstract

“…Previously cited papers in Subsection "Value Estimation" IV-A1 try to first denoise the signal with machine learning or deep learning, while in case of quality detection the goal is to understand whether a certain segment of the input physiological signal is corrupted, and then to proceed with further processing and estimation depending on the evaluation of corruption or by removing from the successive steps the signals too corrupted. We found seven papers [16], [46], [149]- [153] in which the authors perform signal quality classification to improve the performance of vital sign estimation. In two cases [150], [151] the corrupted signals are identified with the help of a clinician, and Shi et al also report a signal quality metric based on the shape of the radar-based heart sound envelope and a threshold based on expert evaluation to provide a signal classification.…”

Section: ) Quality Detectionmentioning

confidence: 99%

“…The other papers consider body movements as artefacts to be recognised by means of a machine learning or deep learning detector. Regarding model choice, we can observe the use of CNN [152], hybrid structure with CNN + SVM [16], Long Short-Term Memory (LSTM) [149], LINEAR machine learning models [150], [151], [153], SVM [46], [150], [153], DT or RF [150], [151], [153], and KNN [150], [153], with accuracies of over 90% in most of the papers. A paper uses the heartbeat from a public ECG dataset to pre-train an LSTM model [149].…”

Section: ) Quality Detectionmentioning

confidence: 99%

Machine Learning in RADAR-based Physiological Signals Sensing: A Scoping Review of the Models, Datasets and Metrics

Nocera,

Senigagliesi,

Raimondi

et al. 2024

Preprint

View full text Add to dashboard Cite

In the field of physiological signals monitoring and its applications, non-contact technology is often proposed as a possible alternative to traditional contact devices. The ability to extract information about a patient’s health status in an unobtrusive way, without stressing the subject and without the need of qualified personnel, fuels research in this growing field. Among the various methodologies, RADAR-based non-contact technology is gaining great interest. This scoping review aims to summarize the main research lines concerning RADAR-based physiological sensing and machine learning applications reporting recent trends, issues and gaps with the scientific literature, best methodological practices, employed standards to be followed, challenges, and future directions. After a systematic search and screening, one hundred and ninety two papers were collected following the guidelines of PRISMA (Preferred Reporting Items for Systematic reviews and Meta-Analyses). The included records covered two macro-areas being regression of physiological signals or physiological features (n = 68 papers) and the other a cluster of papers regarding the processing of RADAR-based physiological signals and features applied to four fields of interest, being RADAR-based diagnosis (n = 73), RADAR-based human behaviour monitoring (n = 21), RADAR-based biometrics authentication (n = 18) and RADAR-based affective computing (n = 9). Papers collected under the diagnosis category were further divided, on the basis of their aims: in breath pattern classification (n = 39), infection detection (n = 10), sleep stage classification (n = 9), heart disease detection (n = 8) and quality detection (n = 7). Papers collected under the human behaviour monitoring were further divided based on their aims: fatigue detection (n = 8), human detection (n = 7), human localisation (n = 4), human orientation (n = 2), and activities classification (n = 3).

show abstract

“…This network model speeds up the inference of the model by optimizing the activation function of YOLOv3, but the complexity of the model was still high. Wei and Li et al obtained a lightweight target detection network by replacing the backbone network of YOLOv4 with a lightweight network structure [14][15] . The model complexity is reduced in this way at the expense of the detection accuracy of the model.…”

Section: Introductionmentioning

confidence: 99%

A lightweight vehicles detection network model based on YOLOv5

Xu,

Liu,

Liu

et al. 2023

Fourth International Conference on Artificial Intelligence and Electromechanical Automation (AIEA 2023)

View full text Add to dashboard Cite

For the existing target detection algorithm with large number of parameters and computation, which cannot be broadly applicable in the vehicle detection field, a lightweight vehicle detection network model is designed in this paper. The vehicle detection model mainly uses RFB-PANet to replace the PANet feature fusion network of YOLOv5s to obtain rich feature information to make the model more accurate, and uses GhostConv structure to optimize the traditional convolutional layer of YOLOv5s to have less parameters and computation of the model, so as to achieve the purpose of lightweights vehicle detection model. The experimental results show that proposed vehicle detection network model reduces the number of parameters by 25.0% and the computational effort by 18.75% compared with the original YOLOV5 network model. The mAp is increased by 2.1% and 1.2% on Pascal VOC dataset and vehicle dataset, respectively, thus achieving a good balance of detection accuracy and light weight

show abstract

Small target detection algorithm based on YOLOv4

Cited by 6 publications

References 2 publications

Intellectual Data Aggregation Using Independent Cluster based Medicaid Method for Network Functionality Fabrication

Intellectual Data Aggregation Using Independent Cluster based Medicaid Method for Network Functionality Fabrication

Machine Learning in RADAR-based Physiological Signals Sensing: A Scoping Review of the Models, Datasets and Metrics

A lightweight vehicles detection network model based on YOLOv5

Contact Info

Product

Resources

About