Zhongtian Sun
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Research Interests
Zhongtian Sun has a total of 46 co-authors. Here are the top co-authors by number of shared publications:
- Anoushka Harit
University of Cambridge
8 shared publications - Jongmin Yu
University of Cambridge
5 shared publications - Alexandra Ioana Cristea
Durham University
4 shared publications - Jinhong Yang
Department of Healthcare and Information Technology, Inje University, Gimhae, Republic of Korea
3 shared publications Alexandra I. Cristea
2 shared publicationsAnoushka Harit
University of Cambridge Cambridge United Kingdom
2 shared publications
Latest Publications
- Open Access
- 12 April 2026
- Published by Association for Computing Machinery (ACM)
- p. 1092-1103
Heterophily-Agnostic Hypergraph Neural Networks with Riemannian Local Exchanger
- 12 April 2026
- Published by Association for Computing Machinery (ACM)
- p. 1092-1103
Open AccessHeterophily-Agnostic Hypergraph Neural Networks with Riemannian Local Exchanger
- 1 April 2026
- Published by Springer Nature
- p. 51-62
GLANCE: Graph Logic Attention Network with Cluster Enhancement for Heterophilous Graph Representation Learning
- 1 April 2026
- Published by Springer Nature
- p. 51-62
GLANCE: Graph Logic Attention Network with Cluster Enhancement for Heterophilous Graph Representation Learning
- Open Access
- 11 January 2026
- Published by Springer Nature in Neural Processing Letters
- Vol. 58 (1)
Relation Extraction from the Perspective of the Frequency Domain: Frequency-Domain Aware Gated Graph Attention Network
- 11 January 2026
- Published by Springer Nature in Neural Processing Letters
- Vol. 58 (1)
Open AccessRelation Extraction from the Perspective of the Frequency Domain: Frequency-Domain Aware Gated Graph Attention Network
- Open Access
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 14, 142986-142998
Prototype-Guided Diffusion Model for Multi-Class Unsupervised Anomaly Detection
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 14, 142986-142998
Open AccessPrototype-Guided Diffusion Model for Multi-Class Unsupervised Anomaly Detection
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Neural Networks and Learning Systems
- Vol. PP, 1-14
A Latent Diffusion for Stable Frame Interpolation
This article presents a novel approach to video frame interpolation (VFI), called latent diffusion for stable frame interpolation (LD4SFI). LD4SFI leverages a latent diffusion model (LDM) enhanced by a vector-quantized spatiotemporal variational autoencoder (VQ-STVAE). Our method captures intrinsic orthogonal relationships in high-dimensional spatiotemporal data and seamlessly integrates complementary information across diverse video frame sequences. Given the robust sampling capabilities of LDMs, LD4SFI is conditioned on a disparity map that describes the motion dynamics between two neighboring frames. The disparity map is applied to feed explicit spatiotemporal differences into a diffusion model (DM), thereby improving the spatiotemporal smoothness of the interpolation results. With this, LD4SFI efficiently generates interpolated frames, significantly improving the continuity and visual quality of the video content. On UCF-101, densely annotated video segmentation (DAVIS), and SNU-FILM, LD4SFI achieves competitive or improved accuracy relative to existing state-of-the-art (SOTA) methods. LD4SFI produces 0.016 learned perceptual image patch similarity (LPIPS), 36.219 peak signal-to-noise ratio (PSNR), 0.974 structural similarity index (SSIM), 0.031 FloLPIPS, and 20.105 Fréchet inception distance (FID) for the UCF-101 dataset, and 0.072 LPIPS, 30.261 PSNR, 0.912 SSIM, 0.108 FloLPIPS, and 8.037 FID for the DAVIS dataset. Additionally, LD4SFI achieves the highest measured throughput among DMs on a single V100. Experimental results show that LD4SFI outperforms existing SOTA methods, demonstrating highly competitive SOTA performance across standard VFI benchmarks.
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Neural Networks and Learning Systems
- Vol. PP, 1-14
A Latent Diffusion for Stable Frame Interpolation
This article presents a novel approach to video frame interpolation (VFI), called latent diffusion for stable frame interpolation (LD4SFI). LD4SFI leverages a latent diffusion model (LDM) enhanced by a vector-quantized spatiotemporal variational autoencoder (VQ-STVAE). Our method captures intrinsic orthogonal relationships in high-dimensional spatiotemporal data and seamlessly integrates complementary information across diverse video frame sequences. Given the robust sampling capabilities of LDMs, LD4SFI is conditioned on a disparity map that describes the motion dynamics between two neighboring frames. The disparity map is applied to feed explicit spatiotemporal differences into a diffusion model (DM), thereby improving the spatiotemporal smoothness of the interpolation results. With this, LD4SFI efficiently generates interpolated frames, significantly improving the continuity and visual quality of the video content. On UCF-101, densely annotated video segmentation (DAVIS), and SNU-FILM, LD4SFI achieves competitive or improved accuracy relative to existing state-of-the-art (SOTA) methods. LD4SFI produces 0.016 learned perceptual image patch similarity (LPIPS), 36.219 peak signal-to-noise ratio (PSNR), 0.974 structural similarity index (SSIM), 0.031 FloLPIPS, and 20.105 Fréchet inception distance (FID) for the UCF-101 dataset, and 0.072 LPIPS, 30.261 PSNR, 0.912 SSIM, 0.108 FloLPIPS, and 8.037 FID for the DAVIS dataset. Additionally, LD4SFI achieves the highest measured throughput among DMs on a single V100. Experimental results show that LD4SFI outperforms existing SOTA methods, demonstrating highly competitive SOTA performance across standard VFI benchmarks.
- Open Access
- 14 November 2025
- Published by Association for Computing Machinery (ACM)
- p. 674-682
From News to Returns: A Granger-Causal Hypergraph Transformer on the Sphere
- 14 November 2025
- Published by Association for Computing Machinery (ACM)
- p. 674-682
Open AccessFrom News to Returns: A Granger-Causal Hypergraph Transformer on the Sphere
- Open Access
- 14 November 2025
- Published by Association for Computing Machinery (ACM)
- p. 177-184
Quantifying Semantic Shift in Financial NLP: Robust Metrics for Market Prediction Stability
- 14 November 2025
- Published by Association for Computing Machinery (ACM)
- p. 177-184
Open AccessQuantifying Semantic Shift in Financial NLP: Robust Metrics for Market Prediction Stability
- 1 October 2025
- Published by Zhejiang University Press in Frontiers of Information Technology & Electronic Engineering
- Vol. 26 (10) , 1871-1878
Three trustworthiness challenges in large language model-based financial systems: real-world examples and mitigation strategies
摘要 大语言模型 (LLM) 在金融应用中的集成展现出显著潜力, 可提升决策流程、实现操作自动化并提供个性化服务。 然而, 金融系统的高风险特性要求极高的可信度, 而当前LLM往往难以满足这一要求。 本研究识别并探讨了基于LLM的金融系统中的3大可信度挑战: (1) 逃逸式提示——利用模型对齐漏洞生成有害或违规响应; (2) 幻觉现象——模型产出事实错误的输出误导金融决策; (3) 偏见与公平性问题——LLM内嵌的人口统计或制度偏见可能导致个体或区域遭受不公平对待。 为具体呈现这些风险, 我们设计了3项金融相关测试, 并对涵盖专有与开源家族的主流LLM进行评估。 在所有模型中, 每项测试至少出现一次风险行为。 基于这些发现, 系统性地总结了现有风险缓解策略。 我们认为, 解决这些问题不仅对确保金融领域人工智能的负责任使用至关重要, 更是实现其安全可扩展部署的关键所在。
- 1 October 2025
- Published by Zhejiang University Press in Frontiers of Information Technology & Electronic Engineering
- Vol. 26 (10) , 1871-1878
Three trustworthiness challenges in large language model-based financial systems: real-world examples and mitigation strategies
摘要 大语言模型 (LLM) 在金融应用中的集成展现出显著潜力, 可提升决策流程、实现操作自动化并提供个性化服务。 然而, 金融系统的高风险特性要求极高的可信度, 而当前LLM往往难以满足这一要求。 本研究识别并探讨了基于LLM的金融系统中的3大可信度挑战: (1) 逃逸式提示——利用模型对齐漏洞生成有害或违规响应; (2) 幻觉现象——模型产出事实错误的输出误导金融决策; (3) 偏见与公平性问题——LLM内嵌的人口统计或制度偏见可能导致个体或区域遭受不公平对待。 为具体呈现这些风险, 我们设计了3项金融相关测试, 并对涵盖专有与开源家族的主流LLM进行评估。 在所有模型中, 每项测试至少出现一次风险行为。 基于这些发现, 系统性地总结了现有风险缓解策略。 我们认为, 解决这些问题不仅对确保金融领域人工智能的负责任使用至关重要, 更是实现其安全可扩展部署的关键所在。
- Open Access
EWC-Guided Diffusion Replay for Exemplar-Free Continual Learning in Medical Imaging
Medical imaging foundation models must adapt over time, yet full retraining is often blocked by privacy constraints and cost. We present a continual learning framework that avoids storing patient exemplars by pairing class conditional diffusion replay with Elastic Weight Consolidation. Using a compact Vision Transformer backbone, we evaluate across eight MedMNIST v2 tasks and CheXpert. On CheXpert our approach attains 0.851 AUROC, reduces forgetting by more than 30\% relative to DER\texttt{++}, and approaches joint training at 0.869 AUROC, while remaining efficient and privacy preserving. Analyses connect forgetting to two measurable factors: fidelity of replay and Fisher weighted parameter drift, highlighting the complementary roles of replay diffusion and synaptic stability. The results indicate a practical route for scalable, privacy aware continual adaptation of clinical imaging models.
Open AccessEWC-Guided Diffusion Replay for Exemplar-Free Continual Learning in Medical Imaging
Medical imaging foundation models must adapt over time, yet full retraining is often blocked by privacy constraints and cost. We present a continual learning framework that avoids storing patient exemplars by pairing class conditional diffusion replay with Elastic Weight Consolidation. Using a compact Vision Transformer backbone, we evaluate across eight MedMNIST v2 tasks and CheXpert. On CheXpert our approach attains 0.851 AUROC, reduces forgetting by more than 30\% relative to DER\texttt{++}, and approaches joint training at 0.869 AUROC, while remaining efficient and privacy preserving. Analyses connect forgetting to two measurable factors: fidelity of replay and Fisher weighted parameter drift, highlighting the complementary roles of replay diffusion and synaptic stability. The results indicate a practical route for scalable, privacy aware continual adaptation of clinical imaging models.
- Open Access
- 22 September 2025
- Published by Association for Computing Machinery (ACM)
- p. 1284-1289
RicciFlowRec: A Geometric Root Cause Recommender Using Ricci Curvature on Financial Graphs
- 22 September 2025
- Published by Association for Computing Machinery (ACM)
- p. 1284-1289
Open AccessRicciFlowRec: A Geometric Root Cause Recommender Using Ricci Curvature on Financial Graphs