Jiangtao Cui

ORCID: 0000-0001-5569-0780
Publications
Citations
Views
---
Saved
---
About
Contact & Profiles
Research Areas
  • Data Management and Algorithms
  • Advanced Image and Video Retrieval Techniques
  • Image Retrieval and Classification Techniques
  • Privacy-Preserving Technologies in Data
  • Cryptography and Data Security
  • Complex Network Analysis Techniques
  • Opinion Dynamics and Social Influence
  • Video Analysis and Summarization
  • Human Mobility and Location-Based Analysis
  • Caching and Content Delivery
  • Advanced Graph Neural Networks
  • Graph Theory and Algorithms
  • Data Mining Algorithms and Applications
  • Traffic Prediction and Management Techniques
  • Social Media and Politics
  • Cloud Computing and Resource Management
  • Time Series Analysis and Forecasting
  • Smart Parking Systems Research
  • Medical Image Segmentation Techniques
  • Anomaly Detection Techniques and Applications
  • Video Surveillance and Tracking Methods
  • Advanced Database Systems and Queries
  • Automated Road and Building Extraction
  • Complexity and Algorithms in Graphs
  • SARS-CoV-2 and COVID-19 Research

Xidian University
2016-2025

Wuhan University
2022

Qingdao University
2021

Institute of Computing Technology
2021

Harbin University of Science and Technology
2018

Sun Yat-sen University
2018

PRG S&Tech (South Korea)
2017

Nanjing University of Aeronautics and Astronautics
2006-2008

Nagoya Institute of Technology
2002-2005

For accommodating rapidly increasing power demands, systems are transitioning from analog to with digital control and communications. Although this modernization brings many far-reaching benefits, the hardware software newly incorporated into also incur vulnerabilities. By taking advantage of these vulnerabilities, adversaries can launch various cyber/physical attacks tamper electricity meter readings, i.e., steal electricity. It is reported that total worldwide annual economic losses caused...

10.1109/jproc.2021.3139754 article EN Proceedings of the IEEE 2022-01-19

Hash-based methods achieve fast similarity search by representing high-dimensional data with compact binary codes. However, both generating codes and encoding unseen effectively efficiently remain very challenging tasks. In this article, we focus on these tasks to implement approximate proposing a novel hash based method named sparse hashing (SH for short). To generate interpretable (or semantically meaningful) codes, the proposed SH first converts original into low-dimensional through...

10.1145/2457465.2457469 article EN ACM transactions on office information systems 2013-05-01

Slit-lamp images play an essential role for diagnosis of pediatric cataracts. We present a computer vision-based framework the automatic localization and slit-lamp by identifying lens region interest (ROI) employing deep learning convolutional neural network (CNN). First, three grading degrees are proposed in conjunction with leading ophthalmologists. The ROI is located automated manner original image using two successive applications Candy detection Hough transform, which cropped, resized...

10.1371/journal.pone.0168606 article EN cc-by PLoS ONE 2017-03-17

Public Key Encryption with Keyword Search (PEKS), an indispensable part of searchable encryption, is stock-in-trade for both protecting data and providing operability encrypted data. So far most PEKS schemes have been established on Identity-Based Cryptography (IBC) key escrow problem inherently. Such severely restricts the promotion IBC-based Infrastructure including component. Hence, Certifcateless (CLPKC) efficient to remove such problem. CLPKC introduced into PEKS, a general model...

10.1109/cc.2014.7004528 article EN China Communications 2014-11-01

Graph Convolutional Network (GCN) has achieved extraordinary success in learning effective task-specific representations of nodes graphs. However, regarding Heterogeneous Information (HIN), existing HIN-oriented GCN methods still suffer from two deficiencies: (1) they cannot flexibly explore all possible meta-paths and extract the most useful ones for a target object, which hinders both effectiveness interpretability; (2) often need to generate intermediate meta-path based dense graphs,...

10.1109/tkde.2021.3101356 article EN IEEE Transactions on Knowledge and Data Engineering 2021-01-01

In smart grids, various Internet-of-Things-based (IoT-based) components are massively deployed across the power systems. However, most of these IoT-based have their own vulnerabilities, leveraging which malicious users can launch different cyber/physical attacks to steal electricity. Economic losses caused by electricity theft amount $96 billion in 2017. Most existing detection techniques suffer from either a high deployment cost or low accuracy. To address concerns, we propose novel...

10.1109/tifs.2023.3265884 article EN IEEE Transactions on Information Forensics and Security 2023-01-01

Approximate Nearest Neighbor (ANN) search in high dimensional space has become a fundamental paradigm many applications. Recently, Locality Sensitive Hashing (LSH) and its variants are acknowledged as the most promising solutions to ANN search. However, state-of-the-art LSH approaches suffer from drawback: accesses candidate objects require large number of random I/O operations. In order guarantee quality returned results, sufficient should be verified, which would consume enormous cost. To...

10.14778/2732939.2732947 article EN Proceedings of the VLDB Endowment 2014-05-01

Since its introduction in 2003, the influence maximization (IM) problem has drawn significant research attention literature. The aim of IM, which is NP-hard, to select a set <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$k$ </tex-math></inline-formula> users known as seed who can most individuals social network. state-of-the-art algorithms estimate expected nodes based on sampled diffusion paths. As...

10.1109/tcss.2022.3164667 article EN IEEE Transactions on Computational Social Systems 2022-05-05

10.1016/j.patrec.2005.12.014 article EN Pattern Recognition Letters 2006-03-01

State-of-the-art classical influence maximization (IM) techniques are "competition-unaware" as they assume that a group (company) finds seeds (users) in network independent of other groups who also simultaneously interested finding such the same network. However, reality several often compete for market (e.g., Samsung, HTC, and Apple smart phone market) hence may attempt to select This has led increasing body research devising IM competitive networks. Despite considerable progress made by...

10.1145/2723372.2723710 article EN 2015-05-27

Multi-view data effectively model and characterize the underlying complex systems, multi-view clustering is of great significance for revealing mechanisms which groups objects into different clusters with high intra-cluster low inter-cluster similarity all views. Current algorithms are criticized undesirable performance because they solely focus on either shared features or correlation objects, failing to address heterogeneity structural constraint various To overcome these problems, a novel...

10.1109/tbdata.2021.3128906 article EN IEEE Transactions on Big Data 2021-11-17

Dynamic pricing plays an important role in solving the problems such as traffic load reduction, congestion control, and revenue improvement. Efficient dynamic strategies can increase capacity utilization, total of service providers, satisfaction both passengers drivers. Many proposed technologies focus on short-term optimization face poor scalability modeling long-term goals for limitations solution optimality prohibitive computation. In this article, a deep reinforcement learning framework...

10.1145/3474841 article EN ACM Transactions on Intelligent Systems and Technology 2022-03-03

Traditional approaches to video tagging are designed propagate tags at the same level, such as assigning of training videos (or shots) test shots), generating for when associated with video-level or shot given a collection annotated shots. This paper focuses on automatical video-level. In other words, we aim assign specific from shot. The solves V2S issue by deriving parts in part videos. To achieve goal, first proposes novel Graph Sparse Group Lasso (shorted GSGL) model linearly reconstruct...

10.1109/tmm.2012.2233723 article EN IEEE Transactions on Multimedia 2012-12-12

Ocular images play an essential role in ophthalmological diagnoses. Having imbalanced dataset is inevitable issue automated ocular diseases diagnosis; the scarcity of positive samples always tends to result misdiagnosis severe patients during classification task. Exploring effective computer-aided diagnostic method deal with crucial. In this paper, we develop cost-sensitive deep residual convolutional neural network (CS-ResCNN) classifier diagnose ophthalmic using retro-illumination images....

10.1186/s12938-017-0420-1 article EN cc-by BioMedical Engineering OnLine 2017-11-21

Traffic prediction is the core task of intelligent transportation system (ITS) and accurate traffic can greatly improve utilization public resources. Dynamic interaction multiple spatial relationships will influence accuracy prediction. However, many existing methods only consider static relationships, which restricts To address above problem, in this article, we propose Multi-Graph Fusion Network (DMGF-Net) to model spatial-temporal correlations network. In DMGF-Net, fusion graph designed...

10.1145/3586164 article EN ACM Transactions on Knowledge Discovery from Data 2023-03-03

Ocular images play an essential role in ophthalmology. Current research mainly focuses on computer-aided diagnosis using slit-lamp images, however few studies have been done to predict the progression of ophthalmic disease. Therefore exploring effective approach prediction can help plan treatment strategies and provide early warning for patients. In this study, we present end-to-end temporal sequence network (TempSeq-Net) automatically disease, which includes employing convolutional neural...

10.1371/journal.pone.0201142 article EN cc-by PLoS ONE 2018-07-31

Many big data applications produce a massive amount of high-dimensional, real-time, and evolving streaming data. Clustering such streams with both effectiveness efficiency are critical for these applications. Although there well-known stream clustering algorithms that based on the popular online-offline framework, still face some major challenges. Several questions not answer satisfactorily: How to perform dimensionality reduction effectively efficiently in online dynamic environment? enable...

10.1109/tkde.2020.2990196 article EN IEEE Transactions on Knowledge and Data Engineering 2020-04-23
Coming Soon ...