Hui Zeng
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
Hui Zeng has a total of 39 co-authors. Here are the top co-authors by number of shared publications:
Anjie Peng
Southwest University of Science and Technology
9 shared publications- Anjie Peng
Southwest University of Science and Technology
9 shared publications - Biwei Chen
Belt and Road School, Beijing Normal University, Zhuhai, China, China
5 shared publications Kang Deng
Southwest University of Science and Technology
4 shared publications- Sanshuai Cui
Faculty of Data Science, City University of Macau, Macau, China
4 shared publications Wenxin Yu
Southwest University of Science and Technology
4 shared publications
Latest Publications
- 17 November 2025
- Published by Springer Nature
- p. 208-222
SmoothStormFCA: Convolutional Smoothing Adversarial Camouflage with Complex Weather Robustness
- 17 November 2025
- Published by Springer Nature
- p. 208-222
SmoothStormFCA: Convolutional Smoothing Adversarial Camouflage with Complex Weather Robustness
- Open Access
- 3 July 2025
- Published by MDPI AG in Preprints.org
ViT-StyleGAN2-ADA for Limited-Data Training
Generative Adversarial Networks (GANs) have demonstrated impressive performance in synthesizing high-fidelity images but often suffer from discriminator overfitting when training data is limited. Adaptive Discriminator Augmentation (ADA) overcomes this limitation but fails to preserve global structure. To mitigate this gap, our study integrates a multi-scale Vision Transformer (ViT)-based discriminator within the StyleGAN2-ADA framework; to stabilize training and reduce mode collapse, we utilize global self-attention in the ViT discriminator to model both local texture and global structure. To enhance ADA’s non-leaking stochastic transformations, we incorporate two additional operations: patch dropout and patch shuffle which further diversify the discriminator’s input without compromising the learning signal. Moreover, we incorporate advanced augmentation strategies whose application probabilities are dynamically adjusted based on feedback from the discriminator, thereby ensuring adaptive and effective regularization throughout training. Furthermore, comprehensive modifications to the loss functions, including token-based Path Length Regularization and gradient penalties tailored for the ViT discriminator, are implemented to enhance training stability and convergence. Experimenting with the proposed approach on multiple datasets, we find that it outperforms, or matches, the baseline StyleGAN2-ADA, and other state-of-the-art GANs. These findings position ViT-D-StyleGAN2-ADA as a powerful solution for generative modeling in data-constrained scenarios. Code and models are available at: https://github.com/mahabub657fy3/ViT-D-StyleGAN2-ADA.
- 3 July 2025
- Published by MDPI AG in Preprints.org
Open AccessViT-StyleGAN2-ADA for Limited-Data Training
Generative Adversarial Networks (GANs) have demonstrated impressive performance in synthesizing high-fidelity images but often suffer from discriminator overfitting when training data is limited. Adaptive Discriminator Augmentation (ADA) overcomes this limitation but fails to preserve global structure. To mitigate this gap, our study integrates a multi-scale Vision Transformer (ViT)-based discriminator within the StyleGAN2-ADA framework; to stabilize training and reduce mode collapse, we utilize global self-attention in the ViT discriminator to model both local texture and global structure. To enhance ADA’s non-leaking stochastic transformations, we incorporate two additional operations: patch dropout and patch shuffle which further diversify the discriminator’s input without compromising the learning signal. Moreover, we incorporate advanced augmentation strategies whose application probabilities are dynamically adjusted based on feedback from the discriminator, thereby ensuring adaptive and effective regularization throughout training. Furthermore, comprehensive modifications to the loss functions, including token-based Path Length Regularization and gradient penalties tailored for the ViT discriminator, are implemented to enhance training stability and convergence. Experimenting with the proposed approach on multiple datasets, we find that it outperforms, or matches, the baseline StyleGAN2-ADA, and other state-of-the-art GANs. These findings position ViT-D-StyleGAN2-ADA as a powerful solution for generative modeling in data-constrained scenarios. Code and models are available at: https://github.com/mahabub657fy3/ViT-D-StyleGAN2-ADA.
- Open Access
- 23 April 2025
- Published by MDPI AG in Mathematics
- Vol. 13 (9) , 1372
Hierarchical Feature Fusion and Enhanced Attention Mechanism for Robust GAN-Generated Image Detection
In recent years, with the rapid advancement of deep learning technologies such as generative adversarial networks (GANs), deepfake technology has become increasingly sophisticated. As a result, the generated fake images are becoming more difficult to visually distinguish from real ones. Existing deepfake detection methods primarily rely on training models with specific datasets. However, these models often suffer from limited generalization when processing images of unknown origin or across domains, leading to a significant decrease in detection accuracy. To address this issue, this paper proposes a deepfake image-detection network based on feature aggregation and enhancement. The key innovation of the proposed method lies in the integration of two modules: the Feature Aggregation Module (FAM) and the Attention Enhancement Module (AEM). The FAM effectively aggregates both deep semantic information and shallow detail features through a multi-scale feature-fusion mechanism, overcoming the limitations of traditional methods that rely on a single-level feature. Meanwhile, the AEM enhances the network’s ability to capture subtle forgery traces by incorporating attention mechanisms and filtering techniques, significantly boosting the model’s efficiency in processing complex information. The experimental results demonstrate that the proposed method achieves significant improvements across all evaluation metrics. Specifically, on the StarGAN dataset, the model attained outstanding performance, with accuracy (Acc) and average precision (AP) both reaching 100%. In cross-dataset testing, the proposed method exhibited strong generalization ability, raising the overall average accuracy to 87.0% and average precision to 92.8%, representing improvements of 5.2% and 6.7%, respectively, compared to existing state-of-the-art methods. These results show that the proposed method can not only achieve optimal performance on data with the same distribution, but also demonstrate strong generalization ability in cross-domain detection tasks.
- 23 April 2025
- Published by MDPI AG in Mathematics
- Vol. 13 (9) , 1372
Open AccessHierarchical Feature Fusion and Enhanced Attention Mechanism for Robust GAN-Generated Image Detection
In recent years, with the rapid advancement of deep learning technologies such as generative adversarial networks (GANs), deepfake technology has become increasingly sophisticated. As a result, the generated fake images are becoming more difficult to visually distinguish from real ones. Existing deepfake detection methods primarily rely on training models with specific datasets. However, these models often suffer from limited generalization when processing images of unknown origin or across domains, leading to a significant decrease in detection accuracy. To address this issue, this paper proposes a deepfake image-detection network based on feature aggregation and enhancement. The key innovation of the proposed method lies in the integration of two modules: the Feature Aggregation Module (FAM) and the Attention Enhancement Module (AEM). The FAM effectively aggregates both deep semantic information and shallow detail features through a multi-scale feature-fusion mechanism, overcoming the limitations of traditional methods that rely on a single-level feature. Meanwhile, the AEM enhances the network’s ability to capture subtle forgery traces by incorporating attention mechanisms and filtering techniques, significantly boosting the model’s efficiency in processing complex information. The experimental results demonstrate that the proposed method achieves significant improvements across all evaluation metrics. Specifically, on the StarGAN dataset, the model attained outstanding performance, with accuracy (Acc) and average precision (AP) both reaching 100%. In cross-dataset testing, the proposed method exhibited strong generalization ability, raising the overall average accuracy to 87.0% and average precision to 92.8%, representing improvements of 5.2% and 6.7%, respectively, compared to existing state-of-the-art methods. These results show that the proposed method can not only achieve optimal performance on data with the same distribution, but also demonstrate strong generalization ability in cross-domain detection tasks.
- Open Access
- 11 April 2025
- Published by Association for the Advancement of Artificial Intelligence (AAAI) in Proceedings of the AAAI Conference on Artificial Intelligence
- Vol. 39 (9) , 9789-9796
Everywhere Attack: Attacking Locally and Globally to Boost Targeted Transferability
Adversarial examples’ (AE) transferability refers to the phenomenon that AEs crafted with one surrogate model can also fool other models. Notwithstanding remarkable progress in untargeted transferability, its targeted counterpart remains challenging. This paper proposes an everywhere scheme to boost targeted transferability. Our idea is to attack a victim image both globally and locally. We aim to optimize ‘an army of targets’ in every local image region instead of the previous works that optimize a high-confidence target in the image. Specifically, we split a victim image into non-overlap blocks and jointly mount a targeted attack on each block. Such a strategy mitigates transfer failures caused by attention inconsistency between surrogate and victim models and thus results in stronger transferability. Our approach is method-agnostic, which means it can be easily combined with existing transferable attacks for even higher transferability. Extensive experiments on ImageNet demonstrate that the proposed approach universally improves the state-of-the-art targeted attacks by a clear margin, e.g., the transferability of the widely adopted Logit attack can be improved by 28.8%-300%. We also evaluate the crafted AEs on a real-world platform: Google Cloud Vision. Results further support the superiority of the proposed method.
- 11 April 2025
- Published by Association for the Advancement of Artificial Intelligence (AAAI) in Proceedings of the AAAI Conference on Artificial Intelligence
- Vol. 39 (9) , 9789-9796
Open AccessEverywhere Attack: Attacking Locally and Globally to Boost Targeted Transferability
Adversarial examples’ (AE) transferability refers to the phenomenon that AEs crafted with one surrogate model can also fool other models. Notwithstanding remarkable progress in untargeted transferability, its targeted counterpart remains challenging. This paper proposes an everywhere scheme to boost targeted transferability. Our idea is to attack a victim image both globally and locally. We aim to optimize ‘an army of targets’ in every local image region instead of the previous works that optimize a high-confidence target in the image. Specifically, we split a victim image into non-overlap blocks and jointly mount a targeted attack on each block. Such a strategy mitigates transfer failures caused by attention inconsistency between surrogate and victim models and thus results in stronger transferability. Our approach is method-agnostic, which means it can be easily combined with existing transferable attacks for even higher transferability. Extensive experiments on ImageNet demonstrate that the proposed approach universally improves the state-of-the-art targeted attacks by a clear margin, e.g., the transferability of the widely adopted Logit attack can be improved by 28.8%-300%. We also evaluate the crafted AEs on a real-world platform: Google Cloud Vision. Results further support the superiority of the proposed method.
- Open Access
Everywhere Attack: Attacking Locally and Globally to Boost Targeted Transferability
Adversarial examples' (AE) transferability refers to the phenomenon that AEs crafted with one surrogate model can also fool other models. Notwithstanding remarkable progress in untargeted transferability, its targeted counterpart remains challenging. This paper proposes an everywhere scheme to boost targeted transferability. Our idea is to attack a victim image both globally and locally. We aim to optimize 'an army of targets' in every local image region instead of the previous works that optimize a high-confidence target in the image. Specifically, we split a victim image into non-overlap blocks and jointly mount a targeted attack on each block. Such a strategy mitigates transfer failures caused by attention inconsistency between surrogate and victim models and thus results in stronger transferability. Our approach is method-agnostic, which means it can be easily combined with existing transferable attacks for even higher transferability. Extensive experiments on ImageNet demonstrate that the proposed approach universally improves the state-of-the-art targeted attacks by a clear margin, e.g., the transferability of the widely adopted Logit attack can be improved by 28.8%-300%.We also evaluate the crafted AEs on a real-world platform: Google Cloud Vision. Results further support the superiority of the proposed method.
Open AccessEverywhere Attack: Attacking Locally and Globally to Boost Targeted Transferability
Adversarial examples' (AE) transferability refers to the phenomenon that AEs crafted with one surrogate model can also fool other models. Notwithstanding remarkable progress in untargeted transferability, its targeted counterpart remains challenging. This paper proposes an everywhere scheme to boost targeted transferability. Our idea is to attack a victim image both globally and locally. We aim to optimize 'an army of targets' in every local image region instead of the previous works that optimize a high-confidence target in the image. Specifically, we split a victim image into non-overlap blocks and jointly mount a targeted attack on each block. Such a strategy mitigates transfer failures caused by attention inconsistency between surrogate and victim models and thus results in stronger transferability. Our approach is method-agnostic, which means it can be easily combined with existing transferable attacks for even higher transferability. Extensive experiments on ImageNet demonstrate that the proposed approach universally improves the state-of-the-art targeted attacks by a clear margin, e.g., the transferability of the widely adopted Logit attack can be improved by 28.8%-300%.We also evaluate the crafted AEs on a real-world platform: Google Cloud Vision. Results further support the superiority of the proposed method.
- Open Access
Two Heads Are Better Than One: Averaging along Fine-Tuning to Improve Targeted Transferability
With much longer optimization time than that of untargeted attacks notwithstanding, the transferability of targeted attacks is still far from satisfactory. Recent studies reveal that fine-tuning an existing adversarial example (AE) in feature space can efficiently boost its targeted transferability. However, existing fine-tuning schemes only utilize the endpoint and ignore the valuable information in the fine-tuning trajectory. Noting that the vanilla fine-tuning trajectory tends to oscillate around the periphery of a flat region of the loss surface, we propose averaging over the fine-tuning trajectory to pull the crafted AE towards a more centered region. We compare the proposed method with existing fine-tuning schemes by integrating them with state-of-the-art targeted attacks in various attacking scenarios. Experimental results uphold the superiority of the proposed method in boosting targeted transferability. The code is available at github.com/zengh5/Avg_FT.
Open AccessTwo Heads Are Better Than One: Averaging along Fine-Tuning to Improve Targeted Transferability
With much longer optimization time than that of untargeted attacks notwithstanding, the transferability of targeted attacks is still far from satisfactory. Recent studies reveal that fine-tuning an existing adversarial example (AE) in feature space can efficiently boost its targeted transferability. However, existing fine-tuning schemes only utilize the endpoint and ignore the valuable information in the fine-tuning trajectory. Noting that the vanilla fine-tuning trajectory tends to oscillate around the periphery of a flat region of the loss surface, we propose averaging over the fine-tuning trajectory to pull the crafted AE towards a more centered region. We compare the proposed method with existing fine-tuning schemes by integrating them with state-of-the-art targeted attacks in various attacking scenarios. Experimental results uphold the superiority of the proposed method in boosting targeted transferability. The code is available at github.com/zengh5/Avg_FT.
- 4 December 2024
- Published by Springer Nature
- p. 346-359
A Whale Falls, All Thrive: Mitigating Attention Gap to Improve Adversarial Transferability
- 4 December 2024
- Published by Springer Nature
- p. 346-359
A Whale Falls, All Thrive: Mitigating Attention Gap to Improve Adversarial Transferability
- 26 November 2023
- Published by Springer Nature
- p. 558-570
Detecting Adversarial Examples via Classification Difference of a Robust Surrogate Model
- 26 November 2023
- Published by Springer Nature
- p. 558-570
Detecting Adversarial Examples via Classification Difference of a Robust Surrogate Model
- 15 November 2023
- Published by Springer Nature
- p. 563-574
Ignored Details in Eyes: Exposing GAN-Generated Faces by Sclera
- 15 November 2023
- Published by Springer Nature
- p. 563-574
Ignored Details in Eyes: Exposing GAN-Generated Faces by Sclera
- 15 November 2023
- Published by Springer Nature
- p. 172-183
Towards Undetectable Adversarial Examples: A Steganographic Perspective
- 15 November 2023
- Published by Springer Nature
- p. 172-183
Towards Undetectable Adversarial Examples: A Steganographic Perspective