Zi Liang

Zi Liang

Auto-generated profile, based on publicly available information

Zi Liang

Hong Kong Polytechnic University, Kowloon, Hong Kong

Research Focus & Metrics

Academic Career Length

8Years (2019-Present)

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

Representative Publications

View All Publications
28 October 2024
Verified
Conference Paper
Open access
ZZZiyang Zhou
Ziyang Zhou
PWPinghui Wang
Pinghui Wang
ZLZi Liang
Zi Liang
RZRuofei Zhang
Ruofei Zhang
HBHaitao Bai
Haitao Bai
Deep neural networks are widely used in retrieval systems. However, they are notoriously vulnerable to attack. Among the various forms of adversarial attacks, the patch attack is one of the most threatening forms. This type of attack can introduce cognitive biases into the retrieval system by inserting deceptive patches into images. Despite the seriousness of this threat, there are still no well-established solutions in image retrieval systems. In this paper, we propose the Pre-denosing Augmented Image Retrieval (PAIR) model, a new approach designed to protect image retrieval systems against adversarial patch attacks. The core strategy of PAIR is to dynamically and randomly reconstruct entire images based on their semantic content. This purifies well-designed patch attacks while preserving the semantic integrity of the images. Furthermore, we present a novel training strategy that incorporates a semantic discriminator. This discriminator significantly improves PAIR's ability to capture real semantics and reconstruct images. Experiments show that PAIR significantly outperforms existing defense methods. It effectively reduces the success rate of two state-of-the-art patch attack methods to below 5%, achieving a 14% improvement over current leading methods. Moreover, in defending against global perturbation attacks, PAIR also achieves competitive results.

2 January 2019 in Mitochondrial DNA Part B
Verified
Journal Article
Open access
XNXiang Nong
Xiang Nong
SZSheng-Nan Zhong
Sheng-Nan Zhong
LCLi Chen
Li Chen
YYYaojun Yang
Yaojun Yang
YXYue Xie
Yue Xie
ZLZi Liang
Zi Liang
Pseudoregma bambucicola is an economically significant bamboo pest which is highly reproductive found in Eastern Asia that influences the normal growth of bamboo, induces the formation of sooty molds, and even causes substantial injury or death among young bamboos. This study provides a theoretical and practical basis for the comprehensive prevention and control of Pseudoregma bambucicola. The results showed that 10 geographical populations of Pseudoregma bambucicola showed high genetic diversity (Hd = 0.879 PI + 0.075, =0.08295 + 0.03901) and the average genetic distance was 0.018. In the haplotype phylogenetic tree constructed, the branches of MP tree and BI tree were basically the same, the Sichuan Basin was clustered into one and the populations from other provinces were clustered into a large branch and had a high posterior probability (the posterior probability of MP tree was 99% and the posterior probability of BI tree was 1). The NJ tree showed the differences among the 10 geographies. It was found that the experimental results were consistent with the actual geographical conditions because different geographical Pseudoregma bambusicola experienced natural isolation due to the possible impact of the natural environment, resulting in no genetic exchange between them, thus having a great genetic variability and high genetic diversity.

15 July 2025 in IEEE Transactions on Dependable and Secure Computing
Verified
Journal Article
Closed access
YXYaxin Xiao
Yaxin Xiao
HHHaibo Hu
Haibo Hu
QYQingqing Ye
Qingqing Ye
LTLi Tang
Li Tang
ZLZi Liang
Zi Liang
HZHuadi Zheng
Huadi Zheng
Model extraction (ME) attacks replicate valuable black-box machine learning (ML) models via malicious query interactions. Cutting-edge attacks focus on actively designing query samples to enhance model fidelity and imprudently adhere to the standard ML training approach. This causes a deviation from the true objective of learning a model over a task. In this paper, we innovatively shift our focus from query selection to training process optimization, aiming to boost the similarity of the copy model with the victim model from neuron to model level. We leverage neuron matching theory to attain this objective and develop a general training booster framework, MEBooster, to fully exploit this theory. MEBooster comprises an initial bootstrapping phase that furnishes initial parameters and an optimal model architecture, followed by a post-processing phase that employs fine-tuning for enhanced neuron matching. Notably, MEBooster can seamlessly integrate with all existing model extraction attacks, enhancing their overall performance. Performance evaluation shows up to 58.10% fidelity gain in image classification. From a defender's perspective, we introduce a novel defensive strategy called Stochastic Norm Enlargement (SNE) to mitigate the risk of such attacks by enlarging the model parameters' norm property in training. Performance evaluation shows up to 58.81% extractability (i.e., fidelity) reduction.

6 March 2026 in ACM Transactions on Asian and Low-Resource Language Information Processing
Verified
Journal Article
Open access
HBHaitao Bai
Haitao Bai
PWPinghui Wang
Pinghui Wang
RZRuofei Zhang
Ruofei Zhang
ZLZi Liang
Zi Liang
ZZZiyang Zhou
Ziyang Zhou
ZSZhou Su
Zhou Su
With the widespread use of social media platforms and people’s increasing dependence on them, social media has emerged as one of the most important channels for advertorials. However, there is currently a lack of research on detecting advertorials on social media platforms. This research focuses on detecting advertorials, a type of advertisement that frequently conceals itself within normal articles, blurring the nature of advertising and deceiving users. To effectively carry out research on advertorial detection, we have constructed a multi-topic advertorial dataset in Chinese with rich social information. This dataset is obtained from the Chinese question-answering platform ZHIHU, and it is publicly available to facilitate further research. 1 Furthermore, we propose AdDetector, a novel dual-tower model that detects advertorials by jointly leveraging the article’s textual and social information. In addition, we use fine-grained sentence-level classification labels to improve the model’s generalization capability on previously unseen topic articles. Experiment results show that our model significantly improves the \(F_1\) score by 1.29% in the intra-domain advertorial detection setting and 1.52% in the transfer setting in comparison with several strong baselines. The extensive ablation studies and thorough performance analyses also validate the complementary and beneficial values of the novel components of AdDetector. We also make our source code publicly available to facilitate future studies. 2 This research provides crucial support for user protection and advertising management.

9 May 2025 in ACM Transactions on Information Systems
Verified
Journal Article
Open access
NXNuo Xu
Nuo Xu
PWPinghui Wang
Pinghui Wang
ZLZi Liang
Zi Liang
JZJunzhou Zhao
Junzhou Zhao
XGXiaohong Guan
Xiaohong Guan
Legal case retrieval aims to automatically scour comparable legal cases based on a given query, which is crucial for offering relevant precedents to support the judgment in intelligent legal systems. Due to similar goals, it is often associated with a similar case matching task. To address them, a daunting challenge is assessing the uniquely defined legal-rational similarity within the judicial domain, which distinctly deviates from the semantic similarities in general text retrieval. Past works either tagged domain-specific factors or incorporated reference laws to capture legal-rational information. However, their heavy reliance on expert or unrealistic assumptions restricts their practical applicability in real-world scenarios. In this article, we propose an end-to-end model named LCM-LAI to solve the above challenges. Through meticulous theoretical analysis, LCM-LAI employs a dependent multi-task learning framework to capture legal-rational information within legal cases by a law article prediction sub-task, without any additional assumptions in inference. In addition, LCM-LAI proposes an article-aware attention mechanism to evaluate the legal-rational similarity between across-case sentences based on the law distribution, which is more effective than semantic similarity. We perform a series of exhaustive experiments that include two different tasks that involving four real-world datasets. The results demonstrate that LCM-LAI achieves state-of-the-art performance.

12 April 2026
Verified
Conference Paper
Open access
ZWZhiyao Wu
Zhiyao Wu
ZLZi Liang
Zi Liang
HHHaibo Hu
Haibo Hu
While Large Language Models (LLMs) are increasingly deployed in Web applications such as search, dialogue, and recommendation systems, their reliance on large-scale Web data raises serious privacy concerns, particularly the risk of memorizing sensitive content. Existing Membership Inference Attacks (MIA) rely heavily on the surface form of inputs, rendering them ineffective against semantically preserved but structurally altered samples. This methodological weakness results in widespread false negatives and compromises the integrity of privacy evaluations in large-scale Web corpora. To address this limitation, we propose Adversarial Semantic Membership Inference Attack (ASMIA). ASMIA enhances MIA effectiveness by generating semantically diverse adversarial samples, extracting multi-layer attention features from the target model, and training a contrastive classifier that leverages similarity metrics and logarithmic probabilities to distinguish members from non-members. Experiments on LLMs trained with Wikipedia, a representative large-scale Web corpus, demonstrate that ASMIA significantly outperforms existing methods, highlighting the value of semantic perturbations and attention patterns in detecting training data leakage.