Mark Junjie Li

Dr. Mark Junjie Li


Research Focus & Metrics

Academic Career Length

12Years (2015-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
1 February 2020 in Applied Soft Computing
Verified
Journal Article
Closed access
Muhammad AzharMuhammad Azhar
Muhammad Azhar
JHJoshua Zhexue Huang
Joshua Zhexue Huang
AMAbdul Masud
Abdul Masud
MLMark Junjie Li
Mark Junjie Li
LCLaizhong Cui
Laizhong Cui

18 September 2019 in Entropy
Verified
Journal Article
Open access
PDF file
Muhammad AzharMuhammad Azhar
Muhammad Azhar
MLMark Junjie Li
Mark Junjie Li
JHJoshua Zhexue Huang
Joshua Zhexue Huang
Data classification is an important research topic in the field of data mining. With the rapid development in social media sites and IoT devices, data have grown tremendously in volume and complexity, which has resulted in a lot of large and complex high-dimensional data. Classifying such high-dimensional complex data with a large number of classes has been a great challenge for current state-of-the-art methods. This paper presents a novel, hierarchical, gamma mixture model-based unsupervised method for classifying high-dimensional data with a large number of classes. In this method, we first partition the features of the dataset into feature strata by using k-means. Then, a set of subspace data sets is generated from the feature strata by using the stratified subspace sampling method. After that, the GMM Tree algorithm is used to identify the number of clusters and initial clusters in each subspace dataset and passing these initial cluster centers to k-means to generate base subspace clustering results. Then, the subspace clustering result is integrated into an object cluster association (OCA) matrix by using the link-based method. The ensemble clustering result is generated from the OCA matrix by the k-means algorithm with the number of clusters identified by the GMM Tree algorithm. After producing the ensemble clustering result, the dominant class label is assigned to each cluster after computing the purity. A classification is made on the object by computing the distance between the new object and the center of each cluster in the classifier, and the class label of the cluster is assigned to the new object which has the shortest distance. A series of experiments were conducted on twelve synthetic and eight real-world data sets, with different numbers of classes, features, and objects. The experimental results have shown that the new method outperforms other state-of-the-art techniques to classify data in most of the data sets.

16 August 2024 in IEEE Transactions on Multimedia
Verified
Journal Article
Closed access
XWXu Wang
Xu Wang
YLYifan Li
Yifan Li
QZQiudan Zhang
Qiudan Zhang
WWWenhui Wu
Wenhui Wu
MLMark Junjie Li
Mark Junjie Li
JJJianmin Jiang
Jianmin Jiang
Learning to build 3D scene graphs is essential for real-world perception in a structured and rich fashion. However, previous 3D scene graph generation methods utilize a fully supervised learning manner and require a large amount of entity-level annotation data of objects and relations, which is extremely resource-consuming and tedious to obtain. To tackle this problem, we propose 3D-VLAP, a weakly-supervised 3D scene graph generation method via Visual-Linguistic Assisted Pseudo-labeling. Specifically, our 3D-VLAP exploits the superior ability of current large-scale visual-linguistic models to align the semantics between texts and 2D images, as well as the naturally existing correspondences between 2D images and 3D point clouds, and thus implicitly constructs correspondences between texts and 3D point clouds. First, we establish the positional correspondence from 3D point clouds to 2D images via camera intrinsic and extrinsic parameters, thereby achieving alignment of 3D point clouds and 2D images. Subsequently, a large-scale crossmodal visual-linguistic model is employed to indirectly align 3D instances with the textual category labels of objects by matching 2D images with object category labels. The pseudo labels for objects and relations are then produced for 3D-VLAP model training by calculating the similarity between visual embeddings and textual category embeddings of objects and relations encoded by the visual-linguistic model, respectively. Ultimately, we design an edge self-attention based graph neural network to generate scene graphs of 3D point cloud scenes. Extensive experiments demonstrate that our 3D-VLAP achieves comparable results with current advanced fully supervised methods, meanwhile significantly alleviating the pressure of data annotation

9 May 2015
Verified
Book Chapter
Closed access
PDF file
TNThanh-Tung Nguyen
Thanh-Tung Nguyen
HZHe Zhao
He Zhao
JHJoshua Zhexue Huang
Joshua Zhexue Huang
MLMark Junjie Li
Mark Junjie Li

29 March 2018
Verified
Conference Paper
Closed access
SNS. Thomas Ng
S. Thomas Ng
FXFrank J. Xu
Frank J. Xu
YYYifan Yang
Yifan Yang
Hongyang LiHongyang Li
Hongyang Li
MLMark Junjie Li
Mark Junjie Li
Emerging information and communication technologies like building information modeling, internet of things, social networking, cloud computing, big data analytics, mobile application and so on serve as important catalysts to transform the practices of infrastructure asset management (IAM). Owners, occupiers, managers, operators and contractors of infrastructure facilities can now capture, process, share, integrate, and analyze data related to various infrastructure facilities collaboratively and cooperatively to achieve an integrated whole life cycle IAM. This should help improve the overall efficiency of a community and/or city through optimized resource allocation; reduced maintenance and refurbishment cost; less energy consumption; lower risks of service disruption; increased transparency; and better strategic, tactical and operational decisions. This paper outlines the development of a social networking enabled crowdsourcing system for integrated IAM. The system is designed to satisfy the requirements of different stakeholders in managing, operating and repairing a series of interdependent infrastructure facilities. Preliminary studies on the residential sector of Hong Kong reveal that the system could help improve the communication between stakeholders, enhance the transparency and effectiveness of asset management, and reduce the risks of bid-rigging in the repair and maintenance schemes. By integrating the proposed system with the existing systems being adopted throughout the entire building life cycle, such as the building information modeling and asset management systems, facility condition, and usage data that is difficult to obtain through traditional means can now be crowdsourced, co-created, or co-processed for facility performance analysis, complaint analysis, and service scheduling analysis.

7 September 2021
Verified
Book Chapter
Closed access
PDF file
WLWeiqin Lin
Weiqin Lin
XBXianyu Bao
Xianyu Bao
MLMark Junjie Li
Mark Junjie Li