A5100372090- Optical measurement and interference techniques
- Advanced Vision and Imaging
- 3D Surveying and Cultural Heritage
- Robotics and Sensor-Based Localization
- Image and Video Quality Assessment
- Advanced Image and Video Retrieval Techniques
- Advanced Measurement and Detection Methods
- Advanced Neural Network Applications
- Visual Attention and Saliency Detection
- Image and Signal Denoising Methods
- Video Coding and Compression Technologies
- Structural Health Monitoring Techniques
- earthquake and tectonic studies
- 3D Shape Modeling and Analysis
- Multimedia Communication and Technology
- Multimodal Machine Learning Applications
- Satellite Image Processing and Photogrammetry
- Domain Adaptation and Few-Shot Learning
- Optical Systems and Laser Technology
- Epigenetics and DNA Methylation
- Zebrafish Biomedical Research Applications
- Fault Detection and Control Systems
- Hand Gesture Recognition Systems
- Video Surveillance and Tracking Methods
- UAV Applications and Optimization
Dalian University of Technology
2023-2025
Wuqing District People's Hospital
2024
State Grid Corporation of China (China)
2024
Wuhan University of Science and Technology
2022-2024
Anhui University of Technology
2021-2024
Changchun University of Science and Technology
2024
University of Electronic Science and Technology of China
2015-2024
Shanxi University
2024
Shanxi University of Traditional Chinese Medicine
2012-2023
Fuyang Normal University
2023
High-precision geochronologic dating constrains probable causes of Earth's largest mass extinction.
To ensure 360-degree video's continuous playback and reduce the bandwidth waste, predicting user's future fixation is indispensable. However, existing methods concentrate either on motion information or content information. None of them consider users watching behaviors' inconsistency which embodies attention distribution more explicitly. So in this paper, we exploit Cross-Users Behaviors for viewport prediction video adaptive streaming, namely CUB360, trying to concurrently personalized...
This paper tries to address a fundamental question in point cloud self-supervised learning: what is good signal we should leverage learn features from clouds without annotations? To answer that, introduce representation learning framework, based on geometric feature reconstruction. In contrast recent papers that directly adopt masked autoencoder (MAE) and only predict original coordinates or occupancy clouds, our method revisits differences between images identifies three objectives peculiar...
Temperature and load variations significantly affect ultrasonic guided wave propagation therefore lead to the increasing of diagnostic uncertainty wave-based structural health monitoring system, which may cause false reports positives limit practical application method. Most existing researches focus on compensation for a single factor. Therefore it's an important challenge that how compensate waves under influence multiple environmental factors. This study firstly analyzes coupled effect...
Depth sensors like Kinect can provide 2.5D depth data, which shows a extreme difference with regular color data. This paper proposes simple and novel algorithm to perform hand posture classification using sensors. We present the notation of dominant line be used make our descriptor rotation invariant. The recognition system combines SVM classifier achieves robust pose in real time. In cross validation test where half dataset is train while other test, we achieve 97.1% success rate on NTU...
A dynamic binocular stereo vision (DBSV) based on a non-zoom rotating camera in large field of view (FOV) is established herein. novel two-point method proposed to estimate the initial parameters quickly. The intrinsic and roll angle each remains constant, pitch yaw angles after rotates are directly estimated from output high-precision two-axes platform, which makes it possible that three-dimensional (3D) coordinate can be measured online rotation. Moreover, target matching algorithm moving...
Recently, Siamese-based trackers have become the primary research direction in single object tracking. However, a majority of these works aim at improving backbone network tracker to boost its performance, thereby overlooking significant impact that template and search region input on tracking accuracy. To address aforementioned issues, we propose an Asymmetrical Transformer Tracker with Prior Templates (AtptTrack), which includes branch update branch. The function is receive image pairs...
Composite indoor human activity recognition is very important in elderly health monitoring and more difficult than identifying individual movements. This article proposes a sensor-based method that integrates positioning. Convolutional neural networks are used to extract spatial information contained geomagnetic sensors ambient light sensors, while transform encoders temporal motion features collected by gyroscopes accelerometers. We established an model with multimodal feature fusion...
Recent years have witnessed an ever-expanding volume of user-generated content (UGC) videos available on the Internet. Nevertheless, progress perceptual quality assessment UGC still remains quite limited. A distinguished characteristic in complete video production and delivery chain is that they often undergo multiple compression stages before ultimately viewed there does not exist pristine source after are uploaded to hosting platform. To facilitate (VQA), we create a database. It contains...
In this article, we review the general characteristics of seismicity in and around China overall statistics earthquake damage 2021, focusing on several significant events related scientific topics. Among them, largest event is MS 7.4 Madoi Qinghai Province, northwest China. The marks another ≥ 7 occurring near boundary Bayan Har Block that has ended a remarkable quiescence earthquakes within Chinese mainland. addition, 6.4 Yangbi Yunnan southwest draws most attention because its abundant...
Nowadays, binocular stereo vision (BSV) is extensively used in real-time 3D reconstruction, which requires cameras to quickly implement self-calibration. At present, the camera parameters are typically estimated through iterative optimization. The calibration accuracy high, but process time consuming. Hence, a system of BSV with rotating and non-zooming established this study, can rotate horizontally vertically. cameras' intrinsic initial position advance by using Zhang's method. Only yaw...
Abstract Three‐dimensional (3D) documenting and rubbing of rock art is aimed to produce descriptive analytic graphic documentation with metric scale geospatial information for archaeological analysis. Although the integrated surveying technologies aerial photogrammetry, close‐range laser scanning have been widely used generate 3D models sites, implementation these in complex surroundings steep terrain, such as riverside cliffs, remains challenging. In this study, we present an unmanned...
The quantitative mapping of a database that represents finite set classified and/or unclassified data points may be decomposed into three distinctive learning tasks: (1) detection the structure each class model with locally mixture clusters; (2) estimation distributions for induced cluster inside class; and (3) classification classes realizes memberships. function accomplished by probabilistic modular networks then constructed as optimal estimator respect to information theory, tasks can...
Short video applications, like TikTok, have attracted many users across the world. It can feed short videos based on users' preferences and allow to slide boring content anywhere anytime. To reduce loading time keep playback smoothness, most of apps will preload recommended in advance. However, these fixed size order, which lead huge stall bandwidth waste. deal with problems, we present an Adaptive Preloading mechanism for Lyapunov Optimization, also called APL, achieve near-optimal...
Lossy compression introduces complex artifacts such as the blocking, ringing and blurring artifacts, making decoded videos unpleasant for human visual system. In this paper, we propose a coding prior based high efficiency restoration algorithm to remove these artifacts. To improve quality of restored compressed videos, take full advantage side information from streams which is ignored or not fully exploited by most existing post-processing methods. particular, unfiltered frames prediction...
This paper describes the theoretic roadmap of least relative entropy matching two point sets. The novel feature is to align sets without needing establish explicit correspondences. recovery transformational geometry achieved using a mixture principal axes registrations, whose parameters are estimated by minimizing between distributions and expectation-maximization algorithm. We give evidence optimality method we then evaluate algorithm's performance in both rigid nonrigid image registration cases.
The unusual multiwavelength lightcurves of GRB 101225A are revisited by assuming that it is from an off-axis powered a newborn magnetar. We show its optical afterglow lightcurve fitted with the forward shock model parameterizing jet structure as Gaussian function half opening angle core $1.67^{\rm o}$. derived initial Lorentz factor ($\Gamma_0$) 120, and viewing to axis $\theta_v=3.7^{\rm Tentative QPO signatures $P=488$ seconds $P=250\sim 300$ found confidence level 90\% analysing X-ray...