A5101640021- Handwritten Text Recognition Techniques
- Natural Language Processing Techniques
- Topic Modeling
- Optical Wireless Communication Technologies
- Indoor and Outdoor Localization Technologies
- Underwater Vehicles and Communication Systems
- Image Retrieval and Classification Techniques
- Advanced Image Processing Techniques
- Image Enhancement Techniques
- Optical measurement and interference techniques
- Recommender Systems and Techniques
- Image Processing Techniques and Applications
- Acoustic Wave Resonator Technologies
- Cloud Computing and Resource Management
- Image and Signal Denoising Methods
- Distributed and Parallel Computing Systems
- Image Processing and 3D Reconstruction
- Vehicle License Plate Recognition
- Ferroelectric and Piezoelectric Materials
- Advanced Image Fusion Techniques
- Digital and Cyber Forensics
- Advanced Vision and Imaging
- Ultrasound Imaging and Elastography
- Face and Expression Recognition
- Advanced Data Compression Techniques
Jiangxi University of Finance and Economics
2024
Guilin Medical University
2022-2023
South China University of Technology
2018-2022
Shandong University of Science and Technology
2021
Shanghai University
2014-2021
Dongguan University of Technology
2020
Chongqing University
2018-2019
This article presents SVC-onGoing1, an on-going competition for on-line signature verification where researchers can easily benchmark their systems against the state of art in open common platform using large-scale public databases, such as DeepSignDB2 and SVC2021_EvalDB3, standard experimental protocols. SVC-onGoing is based on ICDAR 2021 Competition On-Line Signature Verification (SVC 2021), which has been extended to allow participants anytime. The goal evaluate limits popular scenarios...
Dynamic time warping (DTW) is a popular technique for sequence alignment, and the de facto standard dynamic signature verification. In this paper, we go significant step further to enhance DTW with capability of deep representation learning, propose an end-to-end trainable Deep soft-DTW (DsDTW) model Specifically, design convolutional recurrent adaptive network (CRAN) process signatures, utilize it provide robust discriminative local representations as inputs DTW. As not fully differentiable...
Recently, visible light communication (VLC) has been widely used in indoor positioning. Considering that conventional VLC-based positioning (VLP) algorithms are susceptible to interference and cannot adapt well the complex location environment, we propose a high-precision three-dimensional algorithm based on VLC fingerprinting using fusion of K-means random forest algorithms. Unlike trilateration techniques received signal strength, mentioned system does not depend model parameters is robust...
In this paper, an online to offline (O2O) method based on visible light communication (VLC) is proposed, which different from the traditional VLC with modulation and demodulation. It a new recognition. We use RGB emitting diode (RGB-LED) as transmitter, Pulse Width Modulation (PWM) modulate signal make it flicker at high frequency. Therefore, several features are created. At receiver, complementary metal-oxide-semiconductor (CMOS) image sensor applied our system capture LED images stripes. A...
Indoor visible light positioning has become a research hotspot owing to its high accuracy, low cost, and easy implementation. However, many existing indoor systems based on communication (VLC) can only achieve two-dimensional (2-D) those focusing three-dimensional (3-D) usually contain various sensors or use complex algorithms. Consequently, high-precision 3-D system VLC using modified artificial fish swarm algorithm is proposed, avoiding the problems mentioned above. To simplify solution...
Medical images are usually degraded by numerous noises during acquisition or transmission, which often causes low contrast leading to deterioration of image quality. As such, medical denoising and enhancement has become a paramount routine task. To overcome this problem, we propose cutting-edge joint statistical morphological model for the operation. Firstly, in formulating marginal distribution wavelet coefficients. This is integrated into Bayesian inference framework develop maximum...
In conventional dual-projection moiré topography, the additive-type pattern has an amplitude modulated form, and therefore poor visibility, unable to be improved through low-pass filtering. To overcome this difficulty, paper presents enhanced topography by means of logical moiré. its implementation, a couple projectors project two fringe patterns different colors onto object simultaneously then color camera captures deformed pattern. Instead directly processing fringes, we separate fringes...
With the development of visible light communication, indoor positioning (VLP) becomes popular for researchers in communication industry. However, most existing VLP algorithms only provide solutions on a two-dimensional plane, and those focusing three-dimensional (3-D) usually contain various sensors or hybrid algorithms. To solve these problems, first we transform 3-D location model into an optimization adopt distance based (DBOM) measured distances from terminal to multiple LEDs base...
Although online handwriting verification has made great progress recently, the performances are still far behind real usage owing to small scale of datasets as well limited biometric mediums. Therefore, this paper proposes a new benchmark dataset named Multimodal Signature and Digit String (MSDS), which consists two subsets: MSDS-ChS (Chinese Signatures) MSDS-TDS (Token Strings), contributed by 402 users, with 20 genuine samples skilled forgeries per user subset. handwritten Chinese...
Abstract Medical image is an essential tool used in quantitative and qualitative evaluation of different diseases. imaging methods such as fluorescein angiography (FA), optical coherence tomography (OCTA), computed (CT), (OCT), X-ray are for diagnosis. These modalities suffer from low contrast, which leads to deterioration the quality. Consequently, this causes limitation usage medical images clinical routine hindered its potential by depriving clinicians assessing useful information that...
To address the unsatisfactory recognition of distorted text in images with complicated background, a neural-network-based approach was proposed to detect and recognize text. For detection text, improved CRAFT model applied, deformable convolution introduced replace conventional sufficiently extract features irregular background. On this basis, CRAFT-DCN improve accuracy detection. In order reduce interference on model, separated texts were tailored according coordinates obtained by model....
Clouds provide highly elastic resource provisioning styles through which scientific workflows are allowed to acquire desired resources ahead of the execution and build required software environment on virtual machines (VMs). However, various challenges for cloud workflow, especially its optimal scheduling, yet be addressed. Traditional approaches mainly consider VMs with non-fluctuating, time-invariant, stochastic, or bounded performance. This work describes deployed executed over...
High-dimensional and sparse (HiDS) matrices are the basic inputs of recommender systems. Recently, autoencoder-based approaches to analyzing HiDS from systems becoming increasingly popular, owing their good representative learning ability scalability. However, traditional can usually bring great computational burden on data due its frequent manipulations large latent factor (LF) data. To address this issue, we propose a Fast Autoencoder (FAE)-based recommender. It is capable manipulating...
Low-resolution (LR) scene text images are often encountered in real application scenarios. It is difficult to recognize these texts due the loss of their character details. Super-resolution (SR) an intuitive alternative solve this problem. Unlike generic SR, which mainly focuses on image quality, SR accuracy downstream recognition tasks. Although several methods have been recently proposed, they ignore retaining contours that meaningful for readability and poorly perform long texts....
A series of solid solutions 0.60 BiFeO <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3</sub> -0.35 PbTiO -0.05 Pb(Zn xmlns:xlink="http://www.w3.org/1999/xlink">1/3</sub> Nb xmlns:xlink="http://www.w3.org/1999/xlink">2/3</sub> )O (0.60BF-0.35PT-0.05PZN) were fabricated by state reaction method. This ternary system possessed the coexistence both rhombohedral (R3c) and tetragonal (P4mm) phase. The 0.60BF-0.35PT-0.05PZN ceramics mechanically...
A series of solid solutions 0.60 BiFeO <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3</sub> -0.35 PbTiO -0.05 Pb(Zn xmlns:xlink="http://www.w3.org/1999/xlink">1/3</sub> Nb xmlns:xlink="http://www.w3.org/1999/xlink">2/3</sub> )O (0.60BF-0.35PT-0.05PZN) were fabricated by state reaction method. This ternary system possessed the coexistence both rhombohedral (R3c) and tetragonal (P4mm) phase. The 0.60BF-0.35PT-0.05PZN ceramics mechanically...
This article presents SVC-onGoing, an on-going competition for on-line signature verification where researchers can easily benchmark their systems against the state of art in open common platform using large-scale public databases, such as DeepSignDB and SVC2021_EvalDB, standard experimental protocols. SVC-onGoing is based on ICDAR 2021 Competition On-Line Signature Verification (SVC 2021), which has been extended to allow participants anytime. The goal evaluate limits popular scenarios...