Jiahui Huang
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Jiahui Huang has a total of 27 co-authors. Here are the top co-authors by number of shared publications:
- Yan Pang
Chinese Academy of Sciences
7 shared publications - Changyu Dong
Guangzhou University
3 shared publications - Qiong Wang
Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences
3 shared publications - Sisi Duan
Tsinghua University, Beijing, China
3 shared publications - Teng Huang
Guangzhou University
3 shared publications - Hao Chen
Department of Computer Science and Engineering, The Hong Kong University of Science and Technology, Hong Kong SAR, China
2 shared publications
Latest Publications
- 3 July 2026
- Published by Elsevier in Expert Systems with Applications
- Vol. 332
HiLo: Spatial-spectral hybrid high-low frequency activation for heart and brain vessel segmentation
- 3 July 2026
- Published by Elsevier in Expert Systems with Applications
- Vol. 332
HiLo: Spatial-spectral hybrid high-low frequency activation for heart and brain vessel segmentation
- Open Access
- 21 May 2026
- Published by Springer Nature in International Journal of Educational Technology in Higher Education
- Vol. 23 (1)
LLM-CR: LLM-enhanced course recommendation
Recommender systems have become increasingly important in online learning platforms, where learners face a growing number of available courses and substantial variation in course quality and relevance. However, many existing course recommendation methods primarily rely on user–course interaction sequences and therefore often underutilize rich course semantic information, such as descriptions of learning objectives, covered concepts, and prerequisite-related content. This limitation is particularly problematic in educational settings, where learners’ next-course choices are influenced not only by historical behavior but also by semantic relationships among courses. To address this issue, we propose LLM-CR, a recommendation framework that leverages large language models (LLMs) to enrich course representations with structured semantic information. Specifically, course metadata and textual descriptions are processed offline by an LLM to produce semantically informative course summaries, which are then encoded into dense representations and incorporated into the recommendation pipeline through a lightweight fusion module. In this way, LLM-CR augments conventional behavioral representations with course semantics related to topics, knowledge progression, and prerequisite-relevant information. The resulting model is evaluated in the standard next-course recommendation setting using ranking-based metrics. Experiments on six subject-specific subsets constructed from the XuetangX platform show that LLM-CR consistently improves recommendation performance over strong baseline methods, with especially notable gains on relatively sparse datasets. Additional analyses indicate that the proposed semantic augmentation introduces only modest extra complexity because the most expensive LLM processing is performed offline and reused across training and inference. These results suggest that incorporating LLM-derived semantic features is an effective and practical way to improve course recommendation quality.
- 21 May 2026
- Published by Springer Nature in International Journal of Educational Technology in Higher Education
- Vol. 23 (1)
Open AccessLLM-CR: LLM-enhanced course recommendation
Recommender systems have become increasingly important in online learning platforms, where learners face a growing number of available courses and substantial variation in course quality and relevance. However, many existing course recommendation methods primarily rely on user–course interaction sequences and therefore often underutilize rich course semantic information, such as descriptions of learning objectives, covered concepts, and prerequisite-related content. This limitation is particularly problematic in educational settings, where learners’ next-course choices are influenced not only by historical behavior but also by semantic relationships among courses. To address this issue, we propose LLM-CR, a recommendation framework that leverages large language models (LLMs) to enrich course representations with structured semantic information. Specifically, course metadata and textual descriptions are processed offline by an LLM to produce semantically informative course summaries, which are then encoded into dense representations and incorporated into the recommendation pipeline through a lightweight fusion module. In this way, LLM-CR augments conventional behavioral representations with course semantics related to topics, knowledge progression, and prerequisite-relevant information. The resulting model is evaluated in the standard next-course recommendation setting using ranking-based metrics. Experiments on six subject-specific subsets constructed from the XuetangX platform show that LLM-CR consistently improves recommendation performance over strong baseline methods, with especially notable gains on relatively sparse datasets. Additional analyses indicate that the proposed semantic augmentation introduces only modest extra complexity because the most expensive LLM processing is performed offline and reused across training and inference. These results suggest that incorporating LLM-derived semantic features is an effective and practical way to improve course recommendation quality.
- 14 February 2026
- Published by Elsevier in Expert Systems with Applications
- Vol. 315
OCTMamba: A lightweight ear segmentation framework for 3D portable endoscopic OCT scanner
- 14 February 2026
- Published by Elsevier in Expert Systems with Applications
- Vol. 315
OCTMamba: A lightweight ear segmentation framework for 3D portable endoscopic OCT scanner
- 24 January 2026
- Published by Springer Nature
- p. 502-515
MusoMamba: Mamba-Based Multimodal Ultrasound Image Mutual Learning for Breast Tumor Segmentation
- 24 January 2026
- Published by Springer Nature
- p. 502-515
MusoMamba: Mamba-Based Multimodal Ultrasound Image Mutual Learning for Breast Tumor Segmentation
- Top Cited Papers
- 13 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Medical Imaging
- Vol. 45 (5) , 2360-2372
UltraMamba: Mamba-Based Multimodal Ultrasound Image Adaptive Fusion for Breast Lesion Segmentation
Multimodal ultrasound imaging, combining B-mode ultrasound, shear wave velocity, and shear wave time, is crucial for diagnosing and treating breast lesions, providing insights into lesion characteristics and tissue properties. However, challenges arise from intermodal feature misalignment and attention shifts due to varied capture methods and an overemphasis on vibrant color data. To tackle these issues, we introduce two innovations: a novel segmentation framework and a comprehensive dataset. The UltraMamba framework utilizes bidirectional alignment between modalities and enhances region-specific information to improve breast lesion segmentation accuracy. Key components include the Cross-Modal Knowledge Interaction module for robust information exchange and the Region-Aware Feature Excitation module to focus on relevant features. We also present the BreLS dataset, the first two-dimensional multimodal ultrasound breast lesion dataset, with paired images from 506 cases, serving as a valuable resource for analysis. UltraMamba shows strong performance on the BreLS dataset, achieving a Dice Similarity Coefficient of 72.16% and an HD95 of 42.02 mm, reflecting improvements of 2.59% in DSC and a 6.78 mm reduction in HD95 compared to the second-best framework, MMCA-NET. These results highlight UltraMamba’s potential to enhance segmentation accuracy in clinical settings, facilitating precise treatment planning and, ultimately, leading to improved outcomes. Code: https://github.com/deepang-ai/UltraMamba.
- 13 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Medical Imaging
- Vol. 45 (5) , 2360-2372
Top Cited PapersUltraMamba: Mamba-Based Multimodal Ultrasound Image Adaptive Fusion for Breast Lesion Segmentation
Multimodal ultrasound imaging, combining B-mode ultrasound, shear wave velocity, and shear wave time, is crucial for diagnosing and treating breast lesions, providing insights into lesion characteristics and tissue properties. However, challenges arise from intermodal feature misalignment and attention shifts due to varied capture methods and an overemphasis on vibrant color data. To tackle these issues, we introduce two innovations: a novel segmentation framework and a comprehensive dataset. The UltraMamba framework utilizes bidirectional alignment between modalities and enhances region-specific information to improve breast lesion segmentation accuracy. Key components include the Cross-Modal Knowledge Interaction module for robust information exchange and the Region-Aware Feature Excitation module to focus on relevant features. We also present the BreLS dataset, the first two-dimensional multimodal ultrasound breast lesion dataset, with paired images from 506 cases, serving as a valuable resource for analysis. UltraMamba shows strong performance on the BreLS dataset, achieving a Dice Similarity Coefficient of 72.16% and an HD95 of 42.02 mm, reflecting improvements of 2.59% in DSC and a 6.78 mm reduction in HD95 compared to the second-best framework, MMCA-NET. These results highlight UltraMamba’s potential to enhance segmentation accuracy in clinical settings, facilitating precise treatment planning and, ultimately, leading to improved outcomes. Code: https://github.com/deepang-ai/UltraMamba.
- Open Access
- 30 July 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Big Data
- Vol. 11 (6) , 3356-3368
Graph-Based Contract Sensing Framework for Smart Contract Vulnerability Detection
Smart contract vulnerabilities have led to significant economic losses, threatening blockchain security and development. Graph neural network (GNN)-based approaches, which capture the structural properties of contracts and leverage code dependencies to better understand contract behavior, have become widely used for vulnerability detection. However, these approaches face challenges in losing valuable information during graph construction and failing to capture rich semantic content, while traditional GNNs struggle with long-range dependencies and global context in complex contract graphs. To address these challenges, we propose ConSense, a GNN-based Contract Sensing Framework for Smart Contract Vulnerability Detection. ConSense comprises two core components: the smart contract graph generator, which constructs contract graphs while retaining both structural and semantic information, and ExploreFormer, which effectively integrates local and global context using advanced attention mechanisms for vulnerability detection. Comprehensive experimental evaluations were performed on the IR-ESCD and SCVHunter-SCD datasets. For instance, the IR-ESCD benchmark—which encompasses eight distinct vulnerability categories—demonstrates that ConSense attains an average detection accuracy of 97.74%, with a mean processing time of 0.648 seconds per contract. These results signify a statistically significant improvement over state-of-the-art methods in both precision and computational efficiency.
- 30 July 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Big Data
- Vol. 11 (6) , 3356-3368
Open AccessGraph-Based Contract Sensing Framework for Smart Contract Vulnerability Detection
Smart contract vulnerabilities have led to significant economic losses, threatening blockchain security and development. Graph neural network (GNN)-based approaches, which capture the structural properties of contracts and leverage code dependencies to better understand contract behavior, have become widely used for vulnerability detection. However, these approaches face challenges in losing valuable information during graph construction and failing to capture rich semantic content, while traditional GNNs struggle with long-range dependencies and global context in complex contract graphs. To address these challenges, we propose ConSense, a GNN-based Contract Sensing Framework for Smart Contract Vulnerability Detection. ConSense comprises two core components: the smart contract graph generator, which constructs contract graphs while retaining both structural and semantic information, and ExploreFormer, which effectively integrates local and global context using advanced attention mechanisms for vulnerability detection. Comprehensive experimental evaluations were performed on the IR-ESCD and SCVHunter-SCD datasets. For instance, the IR-ESCD benchmark—which encompasses eight distinct vulnerability categories—demonstrates that ConSense attains an average detection accuracy of 97.74%, with a mean processing time of 0.648 seconds per contract. These results signify a statistically significant improvement over state-of-the-art methods in both precision and computational efficiency.
- 14 July 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Information Forensics and Security
- Vol. 20 (15566013) , 7410-7423
SAMamba: Structure-Aware Mamba for Ethereum Fraud Detection
The pseudonymity nature of Ethereum provides a protective umbrella for criminal activities, allowing criminals to develop a series of black industries such as phishing scams in unregulated areas. In order to exploit the relational inductive bias to discover the real identity of anonymous accounts, graph neural networks (GNNs) have been widely used in Ethereum fraud detection tasks as an effective and powerful framework. However, the expressive power of GNN’s 1-hop message passing mechanism is bounded by the Weisfeiler-Leman (1-WL) test, degrading the fraud detection performance on the Ethereum network. This paper proposes a structure-aware Mamba framework, named SAMamba. Specifically, SAMamba uses a subgraph encoding strategy to capture complex structural patterns and introduces Mamba’s exceptional sequence modeling capabilities to route global information. In order to filter task-relevant information from dense information, the attention mechanism and the selection mechanism are introduced from local and global perspectives, respectively. These tailor-made designs enable SAMamba to distinguish subtle differences in structural patterns and selectively aggregate task-oriented information, thereby demonstrating exceptional performance in fraud detection tasks. Extensive experiments on real-world Ethereum data demonstrate that SAMamba outperforms state-of-the-art methods. The codes are publicly available on Github: https://github.com/deepang-ai/SAMamba.
- 14 July 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Information Forensics and Security
- Vol. 20 (15566013) , 7410-7423
SAMamba: Structure-Aware Mamba for Ethereum Fraud Detection
The pseudonymity nature of Ethereum provides a protective umbrella for criminal activities, allowing criminals to develop a series of black industries such as phishing scams in unregulated areas. In order to exploit the relational inductive bias to discover the real identity of anonymous accounts, graph neural networks (GNNs) have been widely used in Ethereum fraud detection tasks as an effective and powerful framework. However, the expressive power of GNN’s 1-hop message passing mechanism is bounded by the Weisfeiler-Leman (1-WL) test, degrading the fraud detection performance on the Ethereum network. This paper proposes a structure-aware Mamba framework, named SAMamba. Specifically, SAMamba uses a subgraph encoding strategy to capture complex structural patterns and introduces Mamba’s exceptional sequence modeling capabilities to route global information. In order to filter task-relevant information from dense information, the attention mechanism and the selection mechanism are introduced from local and global perspectives, respectively. These tailor-made designs enable SAMamba to distinguish subtle differences in structural patterns and selectively aggregate task-oriented information, thereby demonstrating exceptional performance in fraud detection tasks. Extensive experiments on real-world Ethereum data demonstrate that SAMamba outperforms state-of-the-art methods. The codes are publicly available on Github: https://github.com/deepang-ai/SAMamba.
- Open Access
- 17 June 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Systems, Man, and Cybernetics: Systems
- Vol. 55 (9) , 5839-5852
Hierarchical Network With Local-Global Awareness for Ethereum Account De-anonymization
The expansion of blockchain applications, particularly on platforms like Ethereum, brings escalating security challenges as account anonymity provides breeding grounds for criminals to commit crimes and cause significant economic losses. As the mainstream architecture of de-anonymization technology, graph neural networks (GNNs) provide empirical tools for law enforcement agencies to investigate illegal activities. However, the limited expressiveness of current GNNs leads to performance degradation for Ethereum account de-anonymization. To address this challenge, we propose an innovative Local-Global Awareness (LGA) framework, which consists of a Local Structure-Aware (LSA) module and a Global Information-Aware (GIA) module. LSA integrates subgraph-level encoding strategies with local attention to enhance the capture of microscopic interactions. As a complementary measure, GIA introduces global attention to facilitate the understanding of macroscopic information. The LGA framework meticulously captures subgraph-level account behavior patterns at a granular level while simultaneously incorporating global contextual insights, demonstrating higher-level expressive power and receptive fields over conventional GNN. The efficacy of the LGA framework is corroborated by experimental evaluations conducted on the lw-AIG dataset. Our framework achieves exceptional performance, significantly outstripping state-of-the-art GNN-based methods in terms of the micro F1 score metric, with relative improvements ranging from 0.14% to 6.63%. Through its detailed and comprehensive analysis of account interactions, the LGA framework aims to provide a potent solution to the complex security challenges faced in the expanding blockchain landscape. The code for LGA is available at https://github.com/deepang-ai/LGA.
- 17 June 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Systems, Man, and Cybernetics: Systems
- Vol. 55 (9) , 5839-5852
Open AccessHierarchical Network With Local-Global Awareness for Ethereum Account De-anonymization
The expansion of blockchain applications, particularly on platforms like Ethereum, brings escalating security challenges as account anonymity provides breeding grounds for criminals to commit crimes and cause significant economic losses. As the mainstream architecture of de-anonymization technology, graph neural networks (GNNs) provide empirical tools for law enforcement agencies to investigate illegal activities. However, the limited expressiveness of current GNNs leads to performance degradation for Ethereum account de-anonymization. To address this challenge, we propose an innovative Local-Global Awareness (LGA) framework, which consists of a Local Structure-Aware (LSA) module and a Global Information-Aware (GIA) module. LSA integrates subgraph-level encoding strategies with local attention to enhance the capture of microscopic interactions. As a complementary measure, GIA introduces global attention to facilitate the understanding of macroscopic information. The LGA framework meticulously captures subgraph-level account behavior patterns at a granular level while simultaneously incorporating global contextual insights, demonstrating higher-level expressive power and receptive fields over conventional GNN. The efficacy of the LGA framework is corroborated by experimental evaluations conducted on the lw-AIG dataset. Our framework achieves exceptional performance, significantly outstripping state-of-the-art GNN-based methods in terms of the micro F1 score metric, with relative improvements ranging from 0.14% to 6.63%. Through its detailed and comprehensive analysis of account interactions, the LGA framework aims to provide a potent solution to the complex security challenges faced in the expanding blockchain landscape. The code for LGA is available at https://github.com/deepang-ai/LGA.