A5100359942- Robotic Path Planning Algorithms
- Advanced Vision and Imaging
- Advanced Image Processing Techniques
- Guidance and Control Systems
- Image and Signal Denoising Methods
- Robotics and Sensor-Based Localization
- Distributed Control Multi-Agent Systems
- Image Processing Techniques and Applications
- Optical measurement and interference techniques
- Video Surveillance and Tracking Methods
- Medical Image Segmentation Techniques
- Adaptive Control of Nonlinear Systems
- Aerospace Engineering and Control Systems
- UAV Applications and Optimization
- Computational Geometry and Mesh Generation
- Advanced Neural Network Applications
- Management and Marketing Education
- Digital Media Forensic Detection
- Natural Language Processing Techniques
- Air Traffic Management and Optimization
- Dynamics and Control of Mechanical Systems
- Aeroelasticity and Vibration Control
- Evacuation and Crowd Dynamics
- Infrastructure Maintenance and Monitoring
- Control and Dynamics of Mobile Robots
Chang'an University
2023-2024
Northwestern Polytechnical University
2007-2024
University of Würzburg
2023
East China Normal University
2012-2023
University at Buffalo, State University of New York
2023
Shanghai Key Laboratory of Trustworthy Computing
2023
South China University of Technology
2020
Harbin Institute of Technology
2019
Kunming University of Science and Technology
2018
Shandong University
2014
This paper studies the problem of generating cooperative feasible paths for formation rendezvous unmanned aerial vehicles (UAVs). Cooperative path-planning multi-UAV is mostly a complicated multi-objective optimization with many coupled constraints. In order to satisfy kinematic constraints, i.e., maximum curvature constraint and requirement continuous UAV path, Pythagorean hodograph (PH) curve adopted as parameterized path because its continuity rational intrinsic properties. Inspired by...
This paper introduces a novel benchmark for efficient up-scaling as part of the NTIRE 2023 Real-Time Image Super-Resolution (RTSR) Challenge, which aimed to upscale images from 720p and 1080p resolution native 4K (×2 ×3 factors) in real-time on commercial GPUs. For this, we use new test set containing diverse ranging digital art gaming photography. We assessed methods devised SR by measuring their runtime, parameters, FLOPs, while ensuring minimum PSNR fidelity over Bicubic interpolation....
In recent years, there has been an increasing demand for real-time super-resolution networks on mobile devices. To address this issue, many lightweight models have proposed. However, these still contain time-consuming components that increase inference latency, limiting their real-world applications paper, we propose a novel model single-image based Equivalent Transformation and Dual Stream network construction (ETDS). ET method is proposed to transform operators into time-friendly...
Recent advancements in language-image models have led to the development of highly realistic images that can be generated from textual descriptions. However, increased visual quality these poses a potential threat field media forensics. This paper aims investigate level challenge generation pose To achieve this, we propose new approach leverages DALL-E2 model automatically generate and splice masked regions guided by text prompt. ensure creation manipulations, designed an annotation platform...
An improved artificial potential field (APF) method is proposed for solving local minimum problems. additional force provided and the virtual area developed. The influences of parameters in APF are also discussed genetic algorithm (GA) introduced optimization. Some simulation studies with single obstacle multiple obstacles have carried out to demonstrate effectiveness algorithms.
In this study, we address a coupled task allocation and path planning problem for multi-unmanned aerial vehicle (UAV) reconnaissance/attack. Both are addressed. We propose algorithm maximizing system utility simultaneous arrival. Moreover, consider the targets resource requirement UAVs constraints based on contract net protocol allocation. Benefits of destroying target costs UAV attacking considered in objective function. To multi-UAV problem, combination cooperative particle swarm...
A hierarchical mission planning method was proposed to solve a simultaneous attack problem for multi-unmanned aerial vehicles (UAVs). The consisted of three phases aiming decouple and the problem. In first phase, Pythagorean hodograph (PH) curve used in path estimation process each UAV, which also served as input task allocation process. second algorithm based on negotiation mechanism assign targets. Considering resource requirement, time-dependent value targets consumption UAVs, can...
The coupled task allocation and path planning problem for heterogeneous multiple unmanned aerial vehicles performing a search attack mission involving obstacles no-fly zones are addressed. importance of the target is measured using time-dependent value. A algorithm proposed to obtain maximum system utility. In utility function, reward target, lengths vehicles, number perform simultaneous considered. length based on Pythagorean hodograph curve calculated, it serves as input problem. resource...
Rapidly-exploring Random Tree Star (RRT∗) and its variants can provide a collision-free asymptotic optimal solution for many path planning problems. However, it is inefficient RRT∗ based to rapidly find one initial in clustered environment with narrow passages, since consuming high memory as well time, due large number of iterations sampling critical nodes. To overcome this problem, the paper proposes Locally Guided Multiple Bi-RRT∗ (LGM-BRRT∗) method, which fast by incorporating an improved...
The compound multibody aircraft, created through flexible wingtip docking of multiple unit not only capitalizes on high aerodynamic efficiency, but also holds the potential for in-flight reconfiguration. This study addresses challenges controlling inherently unstable morphing process surfaces, eliminating need additional driving mechanisms and offering advantages in structural simplicity. In-flight controllable was achieved, significantly elevating mission flexibility environmental...
Current research on dynamic displacement measurement based computer vision mostly requires professional high-speed cameras and an ideal shooting environment to ensure the performance accuracy of analysis. However, high cost camera strict requirements sharp image contrast stable during process limit broad application technology. This paper proposes improved method implement multi-point measurements with smartphones in interference environment. A motion-enhanced spatio-temporal context (MSTC)...
Deep convolutional neural networks (CNNs) have achieved unprecedented success in single image super-resolution over the past few years. Meanwhile, there is an increasing demand for with arbitrary scale factors real-world scenarios. Many approaches adopt scale-specific multi-path learning to cope multi-scale a network. However, these methods require large number of parameters. To achieve better balance between reconstruction quality and parameter amounts, we proposes learnable interpolation...
To solve the taxiing control problem of full-wing solar-powered unmanned aerial vehicle (UAV) without front wheel steering servo and rudder, a approach using differential propeller thrust to is proposed in this paper. Firstly, mathematical models two kinds UAVs with wheels turning freely or fixed are established. Meanwhile, characteristics UAV different speeds analyzed. Secondly, based on linear active disturbance rejection (LADRC) theory, yaw angle controller designed by as output. Finally,...
Most postsecondary faculty in the UnitedStates include course goals or objectives as key components of their syllabi. In addition to individual objectives, manyinstitutions have identified institution-wide student learning outcomes(SLOs). This paper describes one facultymembers attempts elicit feedback from students regarding growthrelated both and SLOs provides results acourse evaluation a focus group.
Blind super-resolution, different from conventional non-blind super-resolution based on the assumption of fixed degradation, handles various unknown Gaussian blur kernels, and thus is closer to real-world application. The accuracy kernel estimation deconvolution directly influences performance overall results, but recent works usually introduce artifacts during process. In this paper, we propose our methods a more accurate module (KEM) (DM). Additionally, KEM DM are embedded in structure...
Environmental perception plays a key role in ensuring the operational safety of self-driving vehicles. It involves sensing spatial information such as size, location, orientation car, pedestrian and cyclist surrounding environment. Detection algorithms that fuse LiDAR point clouds images can effectively improve sparsity defect cloud data sensitivity image to harsh environment, but most fusion detection have low accuracy for small objects cyclist. To enhance objects, we propose novel 3D...
Abstract Traffic sign detection is essential to an intelligent driving assistance system. The deep learning‐based algorithm proposed in this paper aims address the issue of low accuracy caused by small size and high density traffic signs real‐world scenarios. First, improve feature extraction module backbone network increase model's ability capture contextual information, partial convolution (PConv) introduced. Second, prevent information loss during downsampling process, a cross‐stage...
In the battle against widespread online misinformation, a growing problem is text-image inconsistency, where images are misleadingly paired with texts different intent or meaning. Existing classification-based methods for inconsistency can identify contextual inconsistencies but fail to provide explainable justifications their decisions that humans understand. Although more nuanced, human evaluation impractical at scale and susceptible errors. To address these limitations, this study...