📝 Publications

Uncovering and Mitigating Destructive Multi-Embedding Attacks in Deepfake Proactive Forensics
L. Jia #, H. Sun #, Z. Guo*, Y. Diao, D. Ma, G. Yang. Proceedings of the AAAI Conference on Artificial Intelligence (AAAI), 2026. (CCF-A) [Code]

• This work identifies the Multi-Embedding Attack (MEA) threat to deepfake watermarks and introduces Adversarial Interference Simulation (AIS), a plug-and-play training paradigm that equips existing methods with robust resistance against such attacks.

WaveGuard: Robust Deepfake Detection and Source Tracing via Dual-Tree Complex Wavelet and Graph Neural Networks
Z. He, Z. Guo*, L. Wang, G. Yang, Y. Diao, and D. Ma*. IEEE Transactions on Circuits and Systems for Video Technology, 2025. (CCF-B, Top Journal) [Code]

• This work proposes a proactive watermarking framework, which explores frequency domain embedding and graph-based structural consistency optimization.

Constructing New Backbone Networks via Space-Frequency Interactive Convolution for Deepfake Detection
Z. Guo, Z. Jia*, L. Wang, D. Wang, G. Yang*, and N. Kasabov. IEEE Transactions on Information Forensics and Security, vol. 19, pp. 401-413, 2024, doi: 10.1109/TIFS.2023.3324739. (CCF-A, Top Journal) [Code]

• This work introduces a novel Space-Frequency Interactive Convolution (SFIConv) that seamlessly upgrades existing backbones to more effectively capture and model deepfake artifacts, significantly boosting detection accuracy and cross-dataset generalization.

LDFnet: Lightweight Dynamic Fusion Network for Face Forgery Detection by Integrating Local Artifacts and Global Texture Information
Z. Guo, L. Wang*, W. Yang, G. Yang*, and K. Li. IEEE Transactions on Circuits and Systems for Video Technology, vol. 34, no. 2, pp. 1255-1265, 2024, doi: 10.1109/TCSVT.2023.3289147.（CCF-B, Top Journal）

• This work achieves a superior accuracy-efficiency trade-off in face forgery detection by effectively integrating local artifact and global texture clues with minimal parameters and computational cost.

Exposing Deepfake Face Forgeries with Guided Residuals
Z. Guo, G. Yang*, J. Chen, and X. Sun. IEEE Transactions on Multimedia, vol. 25, pp. 8458-8470, 2023, doi: 10.1109/TMM.2023.3237169.（CCF-B, Top Journal）

• This work expands the applications of the guided filter, and overcomes the potential bias in the prediction-based residuals.

Beyond Fully Supervised Pixel Annotations: Scribble-Driven Weakly-Supervised Framework for Image Manipulation Localization
S. Li, G. Yu, Z. Guo*, Y. Diao, D. Ma, and G. Yang. Proceedings of the AAAI Conference on Artificial Intelligence (AAAI), 2026. (CCF-A) [Code]

• This study introduces a novel weakly-supervised framework for image manipulation localization using only scribble annotations, which outperforms fully-supervised methods by employing self-supervised consistency, dynamic feature modulation, and entropy minimization.

From Passive Perception to Active Memory: A Weakly Supervised Image Manipulation Localization Framework Driven by Coarse-Grained Annotations
Z. Guo#, D. Xi#, S. Li*, and G. Yang. Proceedings of the AAAI Conference on Artificial Intelligence (AAAI), 2026. (CCF-A) [Code]

• This paper achieves fine-grained image manipulation localization with minimal annotation cost by leveraging coarse box prompts and a subconscious memory-inspired feature fusion strategy.

Similarity Memory Prior is All You Need for Medical Image Segmentation
H. Tang, Z. Guo*, L. Wang*, and C. Liu. IEEE/CVF International Conference on Computer Vision (ICCV), 2025. (CCF-A, Highlight Paper) [Code]

• This paper breaks through the limitations of traditional methods and constructs a category modeling based medical image segmentation network.