Junfeng Zhou

Prof. Junfeng Zhou

School of Computer Science and Technology, Donghua University, Shanghai, China
orcid

Research Focus & Metrics

Academic Career Length

15Years (2012-Present)

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

Representative Publications

View All Publications
26 February 2018 in The VLDB Journal
Verified
Journal Article
Closed access
PDF file
JZJunfeng Zhou
Junfeng Zhou
JYJeffrey Xu Yu
Jeffrey Xu Yu
NLNa Li
Na Li
HWHao Wei
Hao Wei
ZCZiyang Chen
Ziyang Chen
XTXian Tang
Xian Tang

5 August 2024 in Scientific Reports
Verified
Journal Article
Open access
PDF file
TKTwahir Kiobya
Twahir Kiobya
JZJunfeng Zhou
Junfeng Zhou
BMBaraka Maiseli
Baraka Maiseli
MAQBOOL KHANMAQBOOL KHAN
MAQBOOL KHAN
Printed Circuit Boards (PCBs) are key devices for the modern-day electronic technologies. During the production of these boards, defects may occur. Several methods have been proposed to detect PCB defects. However, detecting significantly smaller and visually unrecognizable defects has been a long-standing challenge. The existing two-stage and multi-stage object detectors that use only one layer of the backbone, such as Resnet’s third layer (\(C_4\)) or fourth layer (\(C_5\)), suffer from low accuracy, and those that use multi-layer feature maps extractors, such as Feature Pyramid Network (FPN), incur higher computational cost. Founded by these challenges, we propose a robust, less computationally intensive, and plug-and-play Attentive Context and Semantic Enhancement Module (ACASEM) for two-stage and multi-stage detectors to enhance PCB defects detection. This module consists of two main parts, namely adaptable feature fusion and attention sub-modules. The proposed model, ACASEM, takes in feature maps from different layers of the backbone and fuses them in a way that enriches the resulting feature maps with more context and semantic information. We test our module with state-of-the-art two-stage object detectors, Faster R-CNN and Double-Head R-CNN, and with multi-stage Cascade R-CNN detector on DeepPCB and Augmented PCB Defect datasets. Empirical results demonstrate improvement in the accuracy of defect detection.

21 February 2020 in IEEE Access
Verified
Journal Article
Open access
XTXian Tang
Xian Tang
JZJunfeng Zhou
Junfeng Zhou
YQYaxian Qiu
Yaxian Qiu
XLXiang Liu
Xiang Liu
YSYunyu Shi
Yunyu Shi
JZJingwen Zhao
Jingwen Zhao
Given a directed acyclic graph (DAG) G, G’s transitive reduction (TR) Gtr is the unique DAG satisfying thatGtr has theminimumnumber of edges and has the same transitive closure (TC) asG.TRcomputation has been extensively studied during the past decades andwas used inmany applications, where the main problem is howto computeTRefficiently for large graphs.However, existing approaches have either large space complexity or higher time complexity, whichmakes them cannot computeTRefficiently on large dense graphs. We propose a novel approach forTRcomputation, which takes every single edge as the basic processing unit, and utilizes existing reachability algorithms to test whether it is redundant or not. In thisway,weavoid the costly graph traversal operation of existing approaches.We identify the performance bottleneck and propose a set of heuristics to sort edges, such that to reduce the average processing cost of each edge.We showby experimental results that our approachworksmuchbetter than all the existing approaches, and can be faster than the state-of-the-art approach by morethan two orders of magnitude on large dense graphs.

28 November 2019 in IEEE Access
Verified
Journal Article
Open access
MDMing Du
Ming Du
AYAnping Yang
Anping Yang
JZJunfeng Zhou
Junfeng Zhou
XTXian Tang
Xian Tang
ZCZiyang Chen
Ziyang Chen
YZYanfei Zuo
Yanfei Zuo
Given a directed acyclic graph ( DAG), a k-hop reachability query u?kv is used to answer whether there exists a path from u to v with length ≤k. Answering k-hop reachability queries is a fundamental graph operation and has been extensively studied during the past years. Considering that existing approaches still suffer from inefficiency in practice when processing large graphs, we propose a novel labeling scheme, namely HT, to accelerate k-hop reachability queries answering. HT uses a constrained 2hop distance label to maintain the length of shortest paths between a set of hop nodes and other nodes, and for the remaining reachability information, HT uses a novel topological level to accelerate graph traversal. Further, we propose to enhance HT by two optimization techniques. The experimental results show that compared with the state-of-the-art approaches, HT works best for most graphs when answering k-hop reachability queries with small index size and reasonable index construction time.

22 February 2023 in Applied Sciences
Verified
Journal Article
Open access
PDF file
UMUmar Subhan Malhi
Umar Subhan Malhi
JZJunfeng Zhou
Junfeng Zhou
CYCairong Yan
Cairong Yan
Abdur RasoolAbdur Rasool
Abdur Rasool
SSShahbaz Siddeeq
Shahbaz Siddeeq
MDMing Du
Ming Du
Fashion image clustering is the key to fashion retrieval, forecasting, and recommendation applications. Manual labeling-based clustering is both time-consuming and less accurate. Currently, popular methods for extracting features from data use deep learning techniques, such as a Convolutional Neural Network (CNN). These methods can generate high-dimensional feature vectors, which are effective for image clustering. However, high dimensions can lead to the curse of dimensionality, which makes subsequent clustering difficult. The fashion images-oriented deep clustering method (FIDC) is proposed in this paper. This method uses CNN to generate a 4096-dimensional feature vector for each fashion image through migration learning, then performs dimensionality reduction through a deep-stacked auto-encoder model, and finally performs clustering on these low-dimensional vectors. High-dimensional vectors can represent images, and dimensionality reduction avoids the curse of dimensionality during clustering tasks. A particular point in the method is the joint learning and optimization of the dimensionality reduction process and the clustering task. The optimization process is performed using two algorithms: back-propagation and stochastic gradient descent. The experimental findings show that the proposed method, called FIDC, has achieved state-of-the-art performance.

Verified
Journal Article
Open access
PDF file
UMUmar Subhan Malhi
Umar Subhan Malhi
JZJunfeng Zhou
Junfeng Zhou
Abdur RasoolAbdur Rasool
Abdur Rasool
SSShahbaz Siddeeq
Shahbaz Siddeeq
In fashion e-commerce, predicting item compatibility using visual features remains a significant challenge. Current recommendation systems often struggle to incorporate high-dimensional visual data into graph-based learning models effectively. This limitation presents a substantial opportunity to enhance the precision and effectiveness of fashion recommendations. In this paper, we present the Visual-aware Graph Convolutional Network (VAGCN). This novel framework helps improve how visual features can be incorporated into graph-based learning systems for fashion item compatibility predictions. The VAGCN framework employs a deep-stacked autoencoder to convert the input image’s high-dimensional raw CNN visual features into more manageable low-dimensional representations. In addition to improving feature representation, the GCN can also reason more intelligently about predictions, which would not be possible without this compression. The GCN encoder processes nodes in the graph to capture structural and feature correlation. Following the GCN encoder, the refined embeddings are input to a multi-layer perceptron (MLP) to calculate compatibility scores. The approach extends to using neighborhood information only during the testing phase to help with training efficiency and generalizability in practical scenarios, a key characteristic of our model. By leveraging its ability to capture latent visual features and neighborhood-based learning, VAGCN thoroughly investigates item compatibility across various categories. This method significantly improves predictive accuracy, consistently outperforming existing benchmarks. These contributions tackle significant scalability and computational efficiency challenges, showcasing the potential transformation of recommendation systems through enhanced feature representation, paving the way for further innovations in the fashion domain.