Real-Time Target Detection Method Based on Lightweight Convolutional Neural Network

Frontiers in Bioengineering and Biotechnology, Journal Year: 2022, Volume and Issue: 10

Published: Aug. 16, 2022

The continuous development of deep learning improves target detection technology day by day. current research focuses on improving the accuracy technology, resulting in model being too large. number parameters and speed are very important for practical application embedded systems. This article proposed a real-time method based lightweight convolutional neural network to reduce improve speed. In this article, depthwise separable residual module is constructed combining convolution non-bottleneck-free module, structure used replace VGG backbone SSD feature extraction parameter quantity At same time, kernels 1 × 3 standard adding 1, respectively, obtain multiple graphs corresponding SSD, established integrating information graphs. self-built dataset complex scenes comparative experiments; experimental results verify effectiveness superiority method. tested video performance model, deployed Android platform scalability model.

Language: Английский

---

Improved Multi-Stream Convolutional Block Attention Module for sEMG-Based Gesture Recognition

Frontiers in Bioengineering and Biotechnology, Journal Year: 2022, Volume and Issue: 10

Published: June 7, 2022

As a key technology for the non-invasive human-machine interface that has received much attention in industry and academia, surface EMG (sEMG) signals display great potential advantages field of collaboration. Currently, gesture recognition based on sEMG suffers from inadequate feature extraction, difficulty distinguishing similar gestures, low accuracy multi-gesture recognition. To solve these problems new network called Multi-stream Convolutional Block Attention Module-Gate Recurrent Unit (MCBAM-GRU) is proposed, which signals. The multi-stream formed by embedding GRU module CBAM. Fusing ACC further improves action experimental results show proposed method obtains excellent performance dataset collected this paper with accuracies 94.1%, achieving advanced 89.7% Ninapro DB1 dataset. system high classifying 52 kinds different delay less than 300 ms, showing terms real-time human-computer interaction flexibility manipulator control.

Language: Английский

---

UAV-based Internet of Vehicles: A systematic literature review

Intelligent Systems with Applications, Journal Year: 2023, Volume and Issue: 18, P. 200226 - 200226

Published: April 19, 2023

The Internet of Vehicles (IoV) is a network Internet-enabled intelligent vehicles outfitted with sensors, customized software, and communication technologies, which result the development vehicular ad hoc networks (VANETs). Unmanned aerial (UAVs) have recently sparked much interest in both VANETs IoV ecosystems. UAVs-enabled techniques can be used to facilitate various applications. This paper investigates UAV-based approaches using systematic literature review (SLR) methodology. We 41 research papers on UAV-enabled published from 2018 2022. Our taxonomy reveals that publications are divided into two main categories, including requirements, processing requirements. results our investigation evaluation factors show delay factor has highest frequency 33%, whereas energy consumption 41%. In this paper, we highlight most significantly utilized subjects ecosystem investigate hot keywords, protocols, implementation strategies, performance metrics. Finally, discuss outstanding issues future challenges domain.

Language: Английский

---

Toward Optimization of AGV Path Planning: An RRT^*-ACO Algorithm

IEEE Access, Journal Year: 2024, Volume and Issue: 12, P. 18387 - 18399

Published: Jan. 1, 2024

Automated guided vehicle (AGV) smart parking provides a new solution to solve the problem of urban difficulties. In AGV lot, whether running path is reasonable affects transportation efficiency entire lot. RRT algorithm and ant colony can achieve good planning effect for AGV. However, use uniformly distributed initial pheromones easily lead blind search local optimization in early stage algorithm, resulting decrease efficiency. this paper, an based on fast-scaling random tree-ant (RRT*-ACO) proposed. The method uses fast mechanism RRT* positive feedback advantage combine two algorithms add them algorithm. two-way search, heuristic dynamic sampling step size strategies are used accelerate speed quickly generate path, then deploy different concentrations according guide search. Then, silo excavation strategy added which improves pheromone update rule reduces number iterations Finally, several simulation experiments carried out maps scales, comparative experimental results show that RRT*-ACO has convergence speed, better global ability, significantly higher than traditional system, robustness large-scale maps.

Language: Английский

---

Intelligent planning of fire evacuation routes in buildings based on improved adaptive ant colony algorithm

Computers & Industrial Engineering, Journal Year: 2024, Volume and Issue: 194, P. 110335 - 110335

Published: June 29, 2024

Language: Английский

---

Multi-Objective Location and Mapping Based on Deep Learning and Visual Slam

Sensors, Journal Year: 2022, Volume and Issue: 22(19), P. 7576 - 7576

Published: Oct. 6, 2022

Simultaneous localization and mapping (SLAM) technology can be used to locate build maps in unknown environments, but the constructed often suffer from poor readability interactivity, primary secondary information map cannot accurately grasped. For intelligent robots interact meaningful ways with their environment, they must understand both geometric semantic properties of scene surrounding them. Our proposed method not only reduce absolute positional errors (APE) improve positioning performance system also construct object-oriented dense point cloud output model each object reconstruct indoor scene. In fact, eight categories objects are for detection using coco weights our experiments, most actual reconstructed theory. Experiments show that number points is significantly reduced. The average error Technical University Munich (TUM) datasets very small. camera reduced introduction constraints, improved. At same time, algorithm segment environment high accuracy.

Language: Английский

---

Kinematics characteristics analysis of a 3-UPS/S parallel airborne stabilized platform

Aerospace Science and Technology, Journal Year: 2023, Volume and Issue: 134, P. 108163 - 108163

Published: Feb. 3, 2023

Language: Английский

---

Round-trip multimodal transportation routes planning for foldable vehicle racks

Maritime Policy & Management, Journal Year: 2024, Volume and Issue: unknown, P. 1 - 25

Published: July 30, 2024

Over the past decade, expansion of China's auto exports and pulp imports has presented challenges to international shipping sector. This study examines features inherent in novel paradigm employed by COSCO Shipping Specialized Carriers. Specifically, it involves transportation Foldable Vehicle Racks (FVRs) using multipurpose carriers on outbound trip, conveyance alongside empty FVRs return voyage. Building upon this foundation, we propose a dual-objective route planning model construct round-trip multimodal network. Subsequently, calculate optimal scheme, encompassing routes, changeover nodes, transport modes, hybrid genetic ant colony algorithm. The paper scrutinizes computational results, validates effectiveness algorithm, analyzes scheme variations across different parameter scenarios, along with conducting sensitivity analysis. solution calculated is found reduce transit time 24.77% while improving logistics benefits 4.11% compared firm's existing schedule. Furthermore, explore alternative routes various variation scenarios considering benefit-time preferences. Eventually, six managerial insights are proposed for sustainable development COSCO's FVR business based experimental results.

Language: Английский

---

Target Detection Based on Two-Stream Convolution Neural Network With Self-Powered Sensors Information

IEEE Sensors Journal, Journal Year: 2022, Volume and Issue: 23(18), P. 20681 - 20690

Published: Nov. 16, 2022

With the rapid development of artificial intelligence, a neural network is widely used in various fields. The target detection algorithm mainly based on network, but accuracy greatly related to complexity scene and texture. A RGB-D image from perspective lightweight model integration depth map overcome weak environmental illumination with self-powered sensors information proposed. This article analyzes structure YOLOv4 MobileNet, compares variation parameter numbers between depthwise separable convolution convolutional networks, combines advantages MobileNetv3 network. main three effective feature layers replaced by for initial layer extraction strengthen At same time, standard models are convolution. proposed method compared YOLOv4-MobileNetv3 this article, experimental results show that retains its original accuracy, size about 23% model, processing speed 42% higher than can still reach 83% environment poor lighting conditions.

Language: Английский

---

An optimal method for the Shortest Path Problem: The Dhouib-Matrix-SPP (DM-SPP)

Results in Control and Optimization, Journal Year: 2023, Volume and Issue: 12, P. 100269 - 100269

Published: July 17, 2023

This paper introduces an optimal method entitled Dhouib-Matrix-SPP (DM-SPP) in order to solve the Shortest Path Problem with a complexity time of On+m where n and m are respectively number vertices edges. DM-SPP is rapid method, it can reduce reasonably research consequently running as well energy consumption. It composed five simple steps repeated (n-1) iterations for more clarification guided step-by-step application novel presented. Moreover, several examples solved wherein fully complete random distribute graphs analyzed variation from 20 6000 edges 49 17097565. coded under Python programming language all experimental results demonstrate that rapidly generate short path. Furthermore, by comparing required prerequisite Dijkstra algorithm, be concluded concurrent small instances larger easily outperform especially case incomplete undirected which realistic real-world viewing really not complete. The performance statistically confirmed Mann–Whitney U nonparametric test Further trends will focus on uncertain problem resolution autonomous mobile robot path planning problem.

Language: Английский

---