研究以模式识别与机器学习为主线,二十余年的积累串成一条从数据到模型的完整链条:从多元数据分析与可视化起步,进入对抗样本的生成与防御、协同训练与知识蒸馏,最终走向生成模型与数据增强。对抗样本揭示模型的脆弱之处,生成式数据增强则把这种敏感性转化为可用的多样性——两者指向同一个目标:以更优质的数据与更稳健的学习过程,提升模型的鲁棒性与泛化能力。
Research centres on pattern recognition and machine learning, where two decades of work form a single chain from data to model: multivariate data analysis and visualization first, then the generation and defense of adversarial examples, then collaborative training and knowledge distillation, and finally generative models and data augmentation. Adversarial examples expose where a model is fragile; generative data augmentation turns that same sensitivity into usable diversity — both serving one goal: better data and more robust learning, for models that are accurate, resilient and generalizable.
研究领域为模式识别与机器学习,形成四个相互支撑的板块,时间跨度自 2005 年延续至今:
Research centres on pattern recognition and machine learning, organized into four mutually supporting blocks that span from 2005 to the present:
指纹方向场建模;去噪、重建与检索;基于 GLM 与多种重采样方法的高维离散数据统计分析工具包 mvabund,被收录进 CRAN 国际档案大全。Fingerprint orientation-field modelling; denoising, reconstruction and retrieval; the mvabund toolbox for high-dimensional discrete data built on GLM and multiple resampling methods, archived in CRAN.
对抗机器学习(AML)、对抗训练、对抗检索;生物特征隐私保护;查询安全高效的 CBIR 图像检索系统。Adversarial machine learning (AML), adversarial training and adversarial retrieval; biometric privacy protection; query-secure and query-efficient CBIR systems.
联邦学习、模型蒸馏、数据集蒸馏;师生模型与对抗蒸馏。Federated learning, model distillation and dataset distillation; teacher–student models and adversarial distillation.
数据增强、参数高效微调(PEFT);GAN、自回归 Transformer 与扩散模型驱动的数据生成。Data augmentation and parameter-efficient fine-tuning (PEFT); data generation driven by GANs, autoregressive Transformers and diffusion models.
传统机器学习建立在“静态环境”的平稳性假设之上;真实开放环境打破这一假设,攻击者可通过数据投毒、白盒攻击、黑盒攻击与绕过检测等手段破坏模型。
Classical machine learning rests on a stationarity assumption about a static environment. Real open-world deployment breaks that assumption: attackers can poison data, mount white-box or black-box attacks, or bypass detection.

针对 CBIR 系统提出一种自适应的决策型黑盒算法,大幅减少 87% 的系统查询次数,并在微软必应 Bing 等商业检索引擎上展示了攻击有效性:原始图像的检索结果(如 wardrobe)被对抗样本整体改写为 tapestry、painting、tiles 等无关结果。
For CBIR systems we propose an adaptive decision-based black-box algorithm that cuts system queries by 87%, and demonstrate its effectiveness on commercial engines such as Microsoft Bing: the original retrieval results (e.g. wardrobe) are rewritten by an adversarial sample into unrelated results such as tapestry, painting and tiles.

根据客户端上传到服务器的信息,对其本地数据集实施隐私保护攻击,包括样本重构、参数泄漏、成员推断与属性推断等。
From the information clients upload to the server, we mount privacy attacks on their local datasets — sample reconstruction, parameter leakage, membership inference and property inference.

基模型高层特征的多样化表示可以显著提升集成深度学习模型的对抗稳健性,据此提出优先多样化 Dropout(PDD)与分散式集成梯度(DEG)。
Diversified high-level feature representations across base networks markedly improve the adversarial robustness of deep ensembles — realized as Priority Diversified Dropout (PDD) and Dispersed Ensemble Gradients (DEG).

该方向关注的核心矛盾是:多客户端长尾数据导致协作偏差并影响收敛。研究围绕隐私—效用与通信效率、模型鲁棒性与泛化能力、多模态表征对齐与多路编码器融合、多场景应用下的模型泛化与自适应能力展开。
The central tension: long-tailed data across many clients induces collaboration bias and hurts convergence. Research addresses privacy–utility and communication efficiency, model robustness and generalization, multimodal representation alignment and multi-encoder fusion, and cross-scenario generalization and adaptivity.
多客户端数据呈现长尾分布(Non-IID),带来安全隐私保护、客户端过拟合、收敛慢且不稳定、策略异质性四类挑战;以“联邦蒸馏 + 数据增强”应对:基于实例的自然采样与基于类别的均衡采样协同,在服务器端优化聚合与平滑统计,并对分类器偏差进行校准。
Client data are long-tailed (Non-IID), raising four challenges — privacy protection, client over-fitting, slow and unstable convergence, and strategy heterogeneity. The answer is federated distillation plus data augmentation: instance-based natural sampling combined with class-based balanced sampling, server-side optimization of aggregation and smoothed statistics, and calibration of classifier bias.

通过引入随机变换生成的蒸馏数据,并结合多个子网络的软标签对齐(Logits Transfer),增强模型内部知识多样性,抑制本地过拟合。
Randomly transformed distillation data, combined with soft-label alignment (logits transfer) across multiple sub-networks, strengthens intra-model knowledge diversity and suppresses local over-fitting.

通过类别均衡的第三方无标签代理数据集引导多方位知识迁移,对齐师生模型在其上的特征表示与软标签;三重对齐模型权重、特征表示与软标签输出,学习更紧凑的特征并优化损失空间结构,提升全局模型的鲁棒性和泛化能力。
A class-balanced, third-party unlabeled proxy dataset steers multi-faceted knowledge transfer, aligning teacher and student representations and soft labels on it. Triple alignment — model weights, feature representations and soft-label outputs — yields more compact features and a better-structured loss space, improving global robustness and generalization.

在不同数据异质性指标 α 与长尾分布指标 IF 下,所提方法在 SVHN、CIFAR-10 以及医学影像基准数据集 CheXpert 上均取得更高的全局测试准确率。
Across data-heterogeneity level α and long-tail indicator IF, the method achieves higher global test accuracy on SVHN, CIFAR-10 and the medical imaging benchmark CheXpert.

传统对抗训练在小容量学生模型上提升鲁棒性并不理想,常出现精度与鲁棒性失衡。AdaAD 在对抗博弈框架下优化大小模型之间的蒸馏过程:模拟近似推导二者的最大点对点对齐关系,在内部优化中自适应搜索最优匹配点(锚点),通过小模型生成对抗训练数据以增加大小模型间互动,并设计新的对抗蒸馏目标。
Conventional adversarial training struggles on small-capacity students, often trading accuracy for robustness. AdaAD optimizes teacher–student distillation inside an adversarial-game framework: it derives the maximal point-to-point alignment between the two models, adaptively searches for the best anchor during inner optimization, generates adversarial training data through the student to increase interaction, and designs a new adversarial distillation objective.

涵盖生成式对抗网络 GAN、自回归 Transformer 与 Diffusion 扩散模型三条技术路线,面向四类需求:满足小样本学习、符合常理和常识、减少算力和内存、具备局部编辑可控性——即可靠、鲁棒、高效。
Three technical routes — GANs, autoregressive Transformers and diffusion models — serve four requirements: few-shot capability, commonsense plausibility, reduced compute and memory, and controllable local editing, i.e. reliable, robust and efficient generation.

在主网络中插入鲁棒特征适配器(RFA),通过在特征空间动态自适应生成对抗扰动,有效解决传统对抗训练计算开销大和鲁棒过拟合的问题;推理阶段插拔 RFA 适配器还可用于对抗检测。
A Robustness Feature Adapter (RFA) inserted into the main network dynamically generates adversarial perturbations in feature space, solving the high computational cost and robust over-fitting of conventional adversarial training; the adapter can be plugged in or out at inference for adversarial detection.

从效率与收敛质量两方面评估 RFA:仅更新适配器参数即可显著降低计算开销,尤其适用于大规模预训练模型(如 ViT)的对抗训练;尽管训练速度快,仍能保持较低的 FOSC 值,说明生成的对抗样本质量高,有助于提升鲁棒性泛化能力。该结构亦可迁移至个性化联邦学习,同时增强客户端对抗鲁棒性。
RFA is evaluated on both efficiency and convergence quality: updating only adapter parameters sharply cuts computation, especially for large pre-trained models such as ViT; despite fast training it maintains a low FOSC value, indicating high-quality adversarial examples that improve robustness generalization. The same design transfers to personalized federated learning while strengthening client-side adversarial robustness.

情景自适应的数据增强(IGD)在推理阶段利用外部约束(目标梯度),引导反向扩散过程动态自适应地生成能够等效训练下游任务模型的代理数据集,实现生成模型免微调、高效率地控制反向去噪过程。
Scene-adaptive data augmentation (IGD) uses external constraints — target gradients — at inference time to steer the reverse diffusion process, dynamically generating a proxy dataset that can equivalently train downstream models. The generative model needs no fine-tuning, and the reverse denoising process is controlled efficiently.

核心数据集蒸馏法(Core-Set Selection)将大型真实数据集的核心知识提炼到小规模合成或筛选数据集,可用于持续/增量学习与边缘设备部署;通过 DDPM 部分逆去噪过程中引发的重建偏差,反映并量化潜在的数据似然度。
Core-set selection distils the core knowledge of a large real dataset into a small synthesized or selected set, serving continual/incremental learning and edge deployment. Reconstruction bias induced by partial reverse denoising in DDPM reflects and quantifies the underlying data likelihood.

通过多模态特征融合与分布对齐的协同优化,提高模型对复杂异构数据的表征能力,从而提升整体预测精度。
Coordinated multimodal feature fusion and distribution alignment strengthen the model’s representation of complex heterogeneous data, raising overall predictive accuracy.

数据生成阶段基于对抗博弈的多专家蒸馏融合客户端编码器;模型蒸馏阶段将融合后的集成编码器知识蒸馏到全局模型。由此服务器可通过单轮通信学习全局模型,有效降低客户端的通信成本与安全风险。
In the data-generation stage, multi-expert distillation under an adversarial game fuses client encoders; in the model-distillation stage, the fused ensemble-encoder knowledge is distilled into the global model. The server can therefore learn a global model in a single communication round, cutting client communication cost and security risk.

早期设计依赖人工专业经验与手动设计。超材料具有可编程几何结构,可通过设计几何结构定制特定功能;研究基于体素化表示构建几何候选体结构,并在生成过程中平衡合理性、新颖性与多样性。
Early design relied on expert experience and manual effort. Metamaterials have programmable geometry, so specific functions can be tailored by designing that geometry. Candidates are built on voxelized representations, with generation balancing plausibility, novelty and diversity.

两条路线共享同一套底层机制——学习数据空间中的有效扰动、利用高维空间的“线性敏感性”,最终都服务于模型的鲁棒性与泛化能力;差别在于扰动的目标与约束。
Both routes share the same underlying mechanisms — learning effective perturbations in data space and exploiting the linear sensitivity of high-dimensional space — and both ultimately serve robustness and generalization. They differ in the objective and constraints placed on the perturbation.
| 维度Dimension | 对抗样本Adversarial examples | 生成式数据增强Generative data augmentation |
|---|---|---|
| 扰动目标Perturbation goal | 约束条件下的定向扰动生成,目标“破坏模型性能”Constrained, targeted perturbation aimed at breaking model performance | 分布约束下的自然扰动生成,目标“保持语义、增加变化”Distribution-constrained, natural perturbation aimed at preserving semantics while adding variation |
| 高维线性敏感性Linear sensitivity | 定向累积扰动以实现错误分类Accumulating targeted perturbations to force misclassification | 微小扰动不改变语义,仅增加多样性Small perturbations that leave semantics intact and add diversity |
| 对模型的价值Value to the model | 通过对抗训练暴露模型脆弱点,学会应对攻击(安全威胁)Adversarial training exposes vulnerabilities so the model learns to resist attacks (security threats) | 通过补充训练数据覆盖更多场景,缓解数据稀缺与分布偏移Supplementary training data cover more scenarios, relieving data scarcity and distribution shift |
从对抗样本到生成式数据增强From adversarial examples to generative data augmentation
研究以模式识别与机器学习为主线,自 2005 年延续至今:从多元数据分析与可视化起步,进入对抗样本的生成与防御、协同训练与知识蒸馏,最终走向生成模型与数据增强。对抗样本揭示模型的脆弱之处,生成式数据增强把这种敏感性转化为可用的多样性——两者共同指向同一个目标:以更优质的数据与更稳健的学习过程,提升模型的鲁棒性与泛化能力。
The research follows a single line — pattern recognition and machine learning — from 2005 to the present: it starts with multivariate data analysis and visualisation, moves into the generation and defense of adversarial examples and into collaborative training with knowledge distillation, and arrives at generative models and data augmentation. Adversarial examples expose where a model is fragile; generative data augmentation turns that same sensitivity into usable diversity. Both serve one goal: better data and more robust learning, for models that are accurate, resilient and generalisable.
研究领域为模式识别与机器学习。二十余年的工作形成四个相互支撑的板块,时间跨度自 2005 年延续至今:
The research area is pattern recognition and machine learning. Two decades of work form four mutually supporting blocks, spanning 2005 to the present:
| 板块Block | 核心内容Focus |
|---|---|
| 多元数据分析与可视化Multivariate data analysis & visualisation | 指纹方向场建模;去噪、重建、检索、查询;基于 GLM 与多种重采样方法的高维离散数据统计分析工具包 mvabund,被收录进 CRAN 国际档案大全Fingerprint orientation-field modelling; denoising, reconstruction, retrieval and query; mvabund, a toolbox for high-dimensional discrete data built on GLM with multiple resampling methods, archived in CRAN |
| 对抗样本攻击与防御Attacks and defenses on adversarial examples | 对抗机器学习(AML);生物特征隐私保护;安全高效的 CBIR 图像检索系统Adversarial machine learning (AML); biometric privacy protection; secure and query-efficient CBIR systems |
| 协同训练与知识蒸馏Collaborative training & knowledge distillation | 联邦学习、模型蒸馏、数据集蒸馏Federated learning, model distillation and dataset distillation |
| 生成模型与适配器Generative models & adapters | 数据增强、参数高效微调(PEFT)Data augmentation and parameter-efficient fine-tuning (PEFT) |
该方向覆盖对抗机器学习、对抗训练与对抗检索:从传统学习的平稳性假设出发,给出对抗样本的生成方法、联邦学习中的隐私泄露路径,以及提升集成模型对抗稳健性的途径。
This direction covers adversarial machine learning, adversarial training and adversarial retrieval: starting from the stationarity assumption of classical learning, it presents how adversarial examples are generated, how privacy leaks in federated learning, and how the robustness of deep ensembles can be improved.
静态环境下的平稳性假设是传统机器学习的前提:训练/测试数据独立同分布,统计特性不随时间改变,模型有效性、预测可靠性与统计检验性由此成立。开放环境下,白盒攻击、黑盒攻击、绕过检测与数据投毒使这一假设失效,对抗样本的生成与防御因此成为模型落地的必要环节。
Stationarity in a static environment underpins classical machine learning: training and test data are independent and identically distributed, and statistical properties do not drift, so model validity, prediction reliability and statistical testability hold. In an open environment, white-box attacks, black-box attacks, detection bypassing and data poisoning break that assumption — generating and defending adversarial examples becomes indispensable for deployment.

针对 CBIR 系统提出一种自适应的决策型黑盒算法,大幅减少 87% 的系统查询次数,并成功在微软必应 Bing 等商业检索引擎上展示了攻击有效性:原始检索结果(wardrobe)被对抗检索样本改写为 tapestry、painting、tiles 等无关结果。
For CBIR systems we propose an adaptive decision-based black-box algorithm that cuts system queries by 87%, and we demonstrate its effectiveness on commercial engines such as Microsoft Bing: the original retrieval results (wardrobe) are rewritten by an adversarial retrieval sample into unrelated results such as tapestry, painting and tiles.

根据客户端上传到服务器的信息,对其本地数据集实施隐私保护攻击,包括样本重构、参数泄漏、成员推断、属性推断等。
From the information that clients upload to the server, privacy attacks are mounted on their local datasets — sample reconstruction, parameter leakage, membership inference and property inference.

基模型高层特征的多样化表示可以显著提升集成深度学习模型的对抗稳健性,据此提出优先多样化 Dropout(Priority Diversified Dropout, PDD)与分散式集成梯度(Dispersed Ensemble Gradients, DEG)。
Diversified high-level feature representations across base networks markedly improve the adversarial robustness of deep ensembles, realised as Priority Diversified Dropout (PDD) and Dispersed Ensemble Gradients (DEG).

该方向围绕联邦学习、师生模型、模型蒸馏与对抗蒸馏展开,核心矛盾是:多客户端长尾数据导致协作偏差并影响收敛。研究关注模型鲁棒性与泛化能力、隐私—效用与通信效率、多模态表征对齐与多路编码器融合策略,以及多场景应用下的模型泛化与自适应能力。
This direction covers federated learning, teacher–student models, model distillation and adversarial distillation. The central tension: long-tailed data across many clients induces collaboration bias and hurts convergence. The work addresses model robustness and generalisation, privacy–utility and communication efficiency, multimodal representation alignment with multi-encoder fusion, and cross-scenario generalisation and adaptivity.
多客户端数据呈现长尾分布(Non-IID),带来安全隐私保护、客户端过拟合、收敛慢且不稳定等问题,并造成分类器偏差与泛化能力不足。应对之道是“联邦蒸馏 + 数据增强”:基于实例的自然采样(更适合表示学习)与基于类别的均衡采样(更适合分类器学习)互补,联邦蒸馏优化聚合策略,数据增强平滑统计异质性,并对分类器偏差进行校准。
Client data are long-tailed (non-IID), raising privacy protection, client over-fitting, and slow or unstable convergence, and producing classifier bias with limited generalisation. The answer is federated distillation plus data augmentation: instance-based natural sampling (better suited to representation learning) complements class-based balanced sampling (better suited to classifier learning); federated distillation optimises the aggregation strategy, data augmentation smooths statistical heterogeneity, and classifier bias is calibrated.

通过引入随机变换生成的蒸馏数据,并结合多个子网络的软标签对齐(Logits Transfer),增强模型内部知识多样性,抑制本地过拟合。
Distillation data produced by random transformations, combined with soft-label alignment (logits transfer) across multiple sub-networks, strengthens intra-model knowledge diversity and suppresses local over-fitting.

通过类别均衡的第三方无标签代理数据集引导多方位知识迁移,对齐师生模型在其上的特征表示与软标签。三重对齐模型权重、特征表示与软标签输出,学习更紧凑的特征并优化损失空间结构,提升全局模型的鲁棒性和泛化能力。
A class-balanced, third-party unlabeled proxy dataset steers multi-faceted knowledge transfer, aligning teacher and student representations and soft labels on it. Triple alignment — model weights, feature representations and soft-label outputs — yields more compact features and a better-structured loss space, improving global robustness and generalisation.


传统对抗训练方法在小容量学生模型上提升鲁棒性的效果并不理想,常出现模型精度与鲁棒性之间的失衡。AdaAD 在对抗博弈框架下优化大小模型之间的蒸馏过程:
Conventional adversarial training struggles to improve robustness on small-capacity student models, often trading accuracy for robustness. AdaAD optimises teacher–student distillation inside an adversarial-game framework:

该方向覆盖数据增强、数据集蒸馏与参数高效的模型微调(PEFT):以生成模型为工具,为下游任务合成、筛选或对齐数据。
This direction covers data augmentation, dataset distillation and parameter-efficient fine-tuning (PEFT): generative models are used to synthesise, select or align data for downstream tasks.
图像生成沿三条技术路线展开:生成式对抗网络 GAN、自回归 Transformer 与 Diffusion 扩散模型。四条需求共同定义了好模型的标准——满足小样本学习、符合常理和常识、减少算力和内存、具备局部编辑可控性,即可靠、鲁棒、高效。
Image generation follows three technical routes — generative adversarial networks (GANs), autoregressive Transformers and diffusion models. Four requirements define a good model: few-shot capability, commonsense plausibility, reduced compute and memory, and controllable local editing — that is, reliable, robust and efficient.

在主网络中插入鲁棒特征适配器(RFA),通过在特征空间动态自适应地生成对抗扰动,有效解决传统对抗训练计算开销大和鲁棒过拟合的问题;推理阶段插拔 RFA 适配器还可用于对抗检测。
A Robustness Feature Adapter (RFA) inserted into the main network dynamically generates adversarial perturbations in feature space, solving the high computational cost and robust over-fitting of conventional adversarial training; the adapter can be plugged in or out at inference for adversarial detection.


情景自适应的数据增强(IGD)在推理阶段利用外部约束(目标梯度),引导反扩散过程动态自适应地生成能够等效训练下游任务模型的代理数据集,实现生成模型免微调、高效率地控制反向去噪过程。
Scene-adaptive data augmentation (IGD) uses external constraints — target gradients — at inference time to steer the reverse diffusion process, dynamically generating a proxy dataset that can equivalently train downstream models. The generative model needs no fine-tuning, and the reverse denoising process is controlled efficiently.

核心数据集蒸馏法(Core-Set Selection)将大型真实数据集的核心知识提炼到小规模合成/筛选数据集,可用于持续/增量学习和边缘设备部署;通过 DDPM 部分逆去噪过程中引发的重建偏差,反映并量化潜在的数据似然度。
Core-set selection distils the core knowledge of a large real dataset into a small synthesised or selected set, serving continual/incremental learning and edge deployment. Reconstruction bias induced by partial reverse denoising in DDPM reflects and quantifies the underlying data likelihood.

精准性:通过多模态特征融合与分布对齐的协同优化,提高模型对复杂异构数据的表征能力,从而提升整体预测精度。
Accuracy: coordinated multimodal feature fusion and distribution alignment strengthen the model’s representation of complex heterogeneous data, raising overall predictive accuracy.

允许服务器通过单轮通信学习全局模型,有效降低客户端的通信成本和安全风险;数据生成阶段基于对抗博弈的多专家蒸馏融合客户端编码器,模型蒸馏阶段将融合后的集成编码器知识蒸馏到全局模型。
The server learns a global model through one round of communication, cutting client communication cost and security risk. In the data-generation stage, multi-expert distillation under an adversarial game fuses client encoders; in the model-distillation stage, the fused ensemble-encoder knowledge is distilled into the global model.

设计物理性合理的超材料结构。超材料具有可编程几何结构,可通过设计几何结构定制特定功能;早期设计依赖人工专业经验和手动设计。研究基于体素化表示构建几何候选体结构,生成过程中的关键问题是平衡合理性、新颖性与多样性。
Designing physically plausible metamaterial structures. Metamaterials have programmable geometry, so specific functions can be tailored by designing that geometry; early design relied on expert experience and manual effort. Candidates are built on voxelised representations, and the key issue during generation is balancing plausibility, novelty and diversity.

两条路线共享同一套底层机制——学习数据空间中的有效扰动、利用高维空间的“线性敏感性”,最终都服务于模型的鲁棒性与泛化能力;差别只在于扰动的目标与约束。
Both routes share the same underlying mechanisms — learning effective perturbations in data space and exploiting the linear sensitivity of high-dimensional space — and both ultimately serve robustness and generalisation. They differ only in the objective and the constraints placed on the perturbation.
| 维度Dimension | 对抗样本Adversarial examples | 生成式数据增强Generative data augmentation |
|---|---|---|
| 学习数据空间的有效扰动Effective perturbation in data space | 约束条件下的定向扰动生成,目标“破坏模型性能”Constrained, targeted perturbation generation aimed at breaking model performance | 分布约束下的自然扰动生成,目标“保持语义、增加变化”Distribution-constrained, natural perturbation aimed at preserving semantics while adding variation |
| 利用高维空间的“线性敏感性”Linear sensitivity of high-dimensional space | 定向累积扰动实现错误分类Accumulating targeted perturbations to force misclassification | 微小扰动不改变语义、仅增加多样性Small perturbations that leave semantics intact and add diversity |
| 提升模型的鲁棒性与泛化能力Robustness and generalisation | 通过对抗训练暴露模型的脆弱点,学会应对攻击(安全威胁)Adversarial training exposes vulnerabilities so the model learns to resist attacks (security threats) | 通过补充训练数据覆盖更多场景,缓解数据稀缺、分布偏移Supplementary training data cover more scenarios, relieving data scarcity and distribution shift |
Professor, School of Computer Science and Technology, Dongguan University of Technology
Email: wangyi@dgut.edu.cn
Research Focus: Pattern Recognition, Machine Learning, Adversarial Machine Learning, Federated Learning, and Generative Data Augmentation[cite: 1, 3]
Modern deep learning systems serve as critical infrastructure across computer vision, biometric security, medical imaging, and collaborative computing[cite: 21, 26]. However, deep learning models remain vulnerable to adversarial attacks, privacy leakage, and data distribution shifts[cite: 5, 7, 10, 26].
Our research laboratory focuses on Pattern Recognition and Machine Learning[cite: 3], bridging Adversarial Machine Learning (AML)[cite: 4, 26] and Generative Data Augmentation[cite: 17, 19, 24]. By understanding model vulnerabilities through perturbation design, we build secure, robust, privacy-preserving, and computationally efficient machine learning models[cite: 3, 24, 26].
[ Research Scope ]
│
┌───────────────────────────┼───────────────────────────┐
Adversarial machine learning investigates the trade-off between model utility and security under open-world, static, and dynamic threats[cite: 4, 5, 26].
Collaborative computing faces severe challenges under non-IID (long-tailed) data distributions, client heterogeneity, privacy leakage, and communication bottlenecks[cite: 7, 10, 22].
Generative models (GANs, Diffusion Models, and Auto-regressive Transformers) provide structured, domain-constrained perturbations that solve data scarcity, domain shifts, and privacy constraints[cite: 16, 24].
| Year | Title | Venue / Journal | Key Topic |
|---|---|---|---|
| 2026 | Diffusion Reconstruction-Based Data Likelihood Estimation for Core-Set Selection[cite: 9, 20] | AAAI 2026[cite: 20] | Data Distillation / Diffusion[cite: 20] |
| 2026 | FedMCA: Cross-Attention-Based Personalized Multimodal Federated Learning Under Modality Heterogeneity[cite: 21] | ICMSSP 2026[cite: 21] | Multimodal Medical FL[cite: 21] |
| 2025 | Robustness Feature Adapter for Efficient Adversarial Training[cite: 4, 15, 17] | ECAI 2025[cite: 17] | Efficient Adversarial Training / PEFT[cite: 15, 17] |
| 2025 | Influence-Guided Diffusion for Dataset Distillation[cite: 9, 15, 19] | ICLR 2025[cite: 19] | Generative Dataset Distillation[cite: 19] |
| 2025 | StableGuard: Towards Unified Copyright Protection and Tamper Localisation[cite: 15] | NeurIPS 2025[cite: 15] | AI Safety & Watermarking[cite: 15] |
| 2025 | Optimizing Metamaterial Inverse Design with 3D Conditional Diffusion Model and Data Augmentation[cite: 23] | Adv. Materials Tech.[cite: 23] | Generative Models in AI for Science[cite: 23] |
| 2024 | Federated Learning with Hybrid Knowledge Distillations on Long-Tailed Heterogeneous Client Data[cite: 9, 12, 13] | ECAI 2024[cite: 12, 13] | Long-Tailed Federated Learning[cite: 12, 13] |
| 2023 | Boosting Accuracy and Robustness of Student Models via Adaptive Adversarial Distillation[cite: 4, 9, 14] | IEEE/CVF CVPR 2023[cite: 14] | Adversarial Knowledge Distillation[cite: 14] |
| 2022 | Individual Property Inference Over Collaborative Learning In Deep Feature Space[cite: 7, 9] | IEEE ICME 2022[cite: 7] | Privacy Leakage in FL[cite: 7] |
| 2021 | DAIR: A Query-Efficient Decision-Based Attack On Image Retrieval Systems[cite: 4, 6] | ACM SIGIR 2021[cite: 6] | Black-Box CBIR Attack[cite: 6] |
| 2021 | Adversarial Defence By Diversified Simultaneous Training Of Deep Ensembles[cite: 4, 8] | AAAI 2021[cite: 8] | Deep Ensemble Robustness[cite: 8] |
Prof. Yi Wang received her academic training across top international institutions in Australia, Hong Kong, and Mainland China[cite: 2].
2002–2009 2009–2012 2012–2017 2017–Present Univ. of UNSW Sydney Hong Kong Baptist Univ. Dongguan Univ. of Tech. Melbourne Postdoc / Res. Assoc. Research Assistant Prof. Full Professor (Researcher) (Math & CS) (Computer Science) (Computer CS Dept.)
Our laboratory welcomes self-motivated Master’s students, PhD candidates, and Postdoctoral Researchers interested in Pattern Recognition, Adversarial Safety, Generative AI, and Federated Learning[cite: 1, 3, 25].
基于 Prof. Yi Wang(王艺)研究团队成果整理 | 东莞理工学院 · 计算机科学与技术学院
点击下方折叠区域展开对应语言版本。 Click the sections below to expand the language version you need.
| 中文版 | English Version |
|---|---|
| 👇 展开下方中文版 | 👇 Expand English version below |
现代深度学习系统已成为计算机视觉、生物特征安全、医学影像和协同计算等领域的关键基础设施。然而,深度学习模型仍然面临对抗攻击、隐私泄露和数据分布偏移等严峻挑战。
本研究方向聚焦模式识别与机器学习,贯穿对抗机器学习(AML)与生成式数据增强两大主线,通过理解模型脆弱性来构建安全、鲁棒、隐私保护和计算高效的机器学习模型。
研究范畴
│
┌─────────────────┼─────────────────┐
│ │ │
1.对抗机器学习 2.协同训练与蒸馏 3.生成模型与数据增强
│ │ │
· 攻击与防御 · 联邦学习 · 数据蒸馏
· 检测与鲁棒性 · 知识蒸馏 · PEFT与合成
研究模型在开放环境下的静态与动态威胁,兼顾模型效用与安全性。
| 研究方向 | 核心成果 |
|---|---|
| 高效查询对抗攻击 | 提出 DAIR 算法,针对CBIR图像检索系统的决策型黑盒攻击,系统查询开销降低 87%,成功绕过微软必应等商业检索引擎(ACM SIGIR 2021) |
| 模型无关对抗检测 | 通过对输入空间施加随机扰动,利用输出统计变化模式检测低幅度对抗噪声 |
| 集成特征鲁棒增强 | 提出 PDD 与 DEG 方法,通过多视角基网络特征多样化优化集成模型对抗鲁棒性(AAAI 2021) |
解决非独立同分布(长尾)数据、客户端异构性、隐私泄露与通信瓶颈问题。
| 研究方向 | 核心成果 |
|---|---|
| 长尾异构联邦学习 | FedHyb / FedDDL 框架:客户端实例采样 + 服务器类别均衡代理数据集蒸馏,解决分类器偏差,在极端非独立同分布(α=0.5, IF=100)下表现优异(ECAI 2024) |
| 自适应对抗蒸馏 | AdaAD:在对抗博弈框架下优化小容量学生模型训练,动态搜索对齐锚点,同时提升精度与鲁棒性(CVPR 2023) |
| 隐私推断与泄露审计 | 揭示协同学习中深度特征空间的属性推断、样本重构和成员推断漏洞(IEEE ICME 2022) |
| 单轮通信联邦学习 | SEDAG 框架:基于无数据多专家蒸馏实现单轮服务器聚合,降低通信成本与中毒攻击风险 |
利用 GAN、扩散模型和自回归 Transformer 解决数据稀缺、领域偏移和隐私约束。
| 研究方向 | 核心成果 |
|---|---|
| 鲁棒特征适配器(RFA) | 即插即用适配器,训练中动态生成对抗扰动。ViT对抗训练加速 >4倍,参数开销降至 <1/10,鲁棒性增益 >50%(ECAI 2025) |
| 影响引导扩散(IGD) | 在反向扩散采样中引入目标梯度引导,免训练生成任务自适应代理数据集(ICLR 2025) |
| 扩散重建数据似然估计 | 通过DDPM部分重建偏差量化数据似然度,用于核心集选择,支持增量学习与边缘部署(AAAI 2026) |
| 跨模态医学图像融合 | FedMCA:基于交叉注意力的个性化多模态联邦学习,解决模态异构MRI分割与分类(ICMSSP 2026) |
| 超材料逆向设计 | 3D条件扩散模型 + 拓扑导数数据增强,合成指定弹性模量与电导率的超材料几何结构 |
对抗样本 数据增强
│ │
├── 学习数据空间的有效扰动 ────────┤
│ │
├── 利用高维空间"线性敏感性" ──────┤
│ │
└── 提升模型鲁棒性与泛化能力 ──────┘
| 维度 | 对抗样本 | 数据增强 |
|---|---|---|
| 扰动目标 | 约束条件下定向扰动,目标是"破坏模型性能" | 分布约束下自然扰动,目标是"保持语义、增加变化" |
| 扰动方式 | 定向累积扰动实现错误分类 | 微小扰动不改变语义,仅增加多样性 |
| 核心价值 | 通过对抗训练暴露脆弱点,学会应对安全威胁 | 补充训练数据覆盖更多场景,缓解数据稀缺与分布偏移 |
核心观点: 模型能力上限取决于数据质量、结构与组织方式。
| 年份 | 论文标题 | 发表 venue | 主题 |
|---|---|---|---|
| 2026 | Diffusion Reconstruction-Based Data Likelihood Estimation for Core-Set Selection | AAAI 2026 | 数据蒸馏/扩散模型 |
| 2026 | FedMCA: Cross-Attention-Based Personalized Multimodal Federated Learning | ICMSSP 2026 | 多模态医学联邦学习 |
| 2025 | Robustness Feature Adapter for Efficient Adversarial Training | ECAI 2025 | 高效对抗训练/PEFT |
| 2025 | Influence-Guided Diffusion for Dataset Distillation | ICLR 2025 | 生成式数据集蒸馏 |
| 2025 | StableGuard: Towards Unified Copyright Protection and Tamper Localisation | NeurIPS 2025 | AI安全与水印 |
| 2024 | Federated Learning with Hybrid Knowledge Distillations on Long-Tailed Data | ECAI 2024 | 长尾联邦学习 |
| 2023 | Boosting Accuracy and Robustness via Adaptive Adversarial Distillation | CVPR 2023 | 对抗知识蒸馏 |
| 2022 | Individual Property Inference Over Collaborative Learning | IEEE ICME 2022 | 联邦学习隐私泄露 |
| 2021 | DAIR: A Query-Efficient Decision-Based Attack On Image Retrieval | ACM SIGIR 2021 | 黑盒CBIR攻击 |
| 2021 | Adversarial Defence By Diversified Simultaneous Training Of Deep Ensembles | AAAI 2021 | 深度集成鲁棒性 |
Modern deep learning systems serve as critical infrastructure across computer vision, biometric security, medical imaging, and collaborative computing. However, deep learning models remain vulnerable to adversarial attacks, privacy leakage, and data distribution shifts.
Our research focuses on Pattern Recognition and Machine Learning, bridging Adversarial Machine Learning (AML) and Generative Data Augmentation. By understanding model vulnerabilities through perturbation design, we build secure, robust, privacy-preserving, and computationally efficient models.
Research Scope
│
┌───────────────────┼───────────────────┐
│ │ │
1.Adversarial ML 2.Collaborative 3.Generative Models
│ Training & Distill. & Data Augment.
│ │ │
· Attack & Defense · Federated Learning · Data Distillation
· Detection & · Knowledge · PEFT & Synthesis
Robustness Distillation
Investigating static and dynamic threats in open-world environments, balancing model utility and security.
| Direction | Key Achievements |
|---|---|
| Query-Efficient Adversarial Attacks | DAIR: a query-efficient decision-based black-box attack for CBIR systems, reducing query overhead by 87% and bypassing Microsoft Bing (ACM SIGIR 2021) |
| Model-Agnostic Adversarial Detection | Detection via random perturbations on input space, leveraging output statistical change patterns to catch low-magnitude adversarial noise |
| Ensemble & Feature Robustness | PDD & DEG: multi-view base-network feature diversification to optimize ensemble robustness against gradient-based attacks (AAAI 2021) |
Addressing non-IID (long-tailed) data, client heterogeneity, privacy leakage, and communication bottlenecks.
| Direction | Key Achievements |
|---|---|
| Long-Tailed Federated Learning | FedHyb / FedDDL: instance-based client sampling + class-balanced proxy distillation at server, resolving classifier bias under extreme non-IID (α=0.5, IF=100) (ECAI 2024) |
| Adaptive Adversarial Distillation | AdaAD: optimizes small-capacity student training via adversarial game framework with dynamic anchor-point search, boosting both accuracy and robustness (CVPR 2023) |
| Privacy Inference & Leakage Auditing | Demonstrated property inference, sample reconstruction, and membership inference in deep feature spaces during collaborative updates (IEEE ICME 2022) |
| One-Shot Federated Learning | SEDAG: single-round server aggregation via data-free multi-expert distillation, cutting communication costs and poisoning exposure |
Leveraging GANs, Diffusion Models, and Auto-regressive Transformers to address data scarcity, domain shifts, and privacy constraints.
| Direction | Key Achievements |
|---|---|
| Robustness Feature Adapter (RFA) | Plug-and-play adapter generating dynamic adversarial perturbations. ViT adversarial training accelerated >4×, parameter overhead reduced to <1/10, robustness gains >50% (ECAI 2025) |
| Influence-Guided Diffusion (IGD) | Target-gradient guidance during reverse diffusion sampling for training-free, task-adaptive proxy dataset generation (ICLR 2025) |
| Diffusion Reconstruction Likelihood | DDPM partial-reconstruction deviations to estimate data likelihood for core-set selection, supporting incremental learning and edge deployment (AAAI 2026) |
| Cross-Modal Medical Image Fusion | FedMCA: cross-attention personalized multimodal FL for modality-heterogeneous MRI segmentation and classification (ICMSSP 2026) |
| Metamaterial Inverse Design | 3D Conditional Diffusion Models + topological derivative data augmentation for synthesizing metamaterial structures with specified elastic modulus and conductivity |
Adversarial Data
Examples Augmentation
│ │
├── Learn effective perturbations in data space ───┤
│ │
├── Exploit "linear sensitivity" in high-dim space ─┤
│ │
└── Enhance model robustness & generalization ──────┘
| Dimension | Adversarial Examples | Data Augmentation |
|---|---|---|
| Perturbation Goal | Directed perturbation under constraints; goal is “to break model performance” | Natural perturbation under distribution constraints; goal is “to preserve semantics, add variation” |
| Perturbation Method | Directional accumulated perturbation causing misclassification | Micro-perturbations that don’t alter semantics, only add diversity |
| Core Value | Exposes vulnerabilities via adversarial training; learns to counter security threats | Supplements training data to cover more scenarios; mitigates data scarcity and distribution shift |
Key Insight: Model capability ceiling depends on data quality, structure, and organization.
| Year | Title | Venue | Topic |
|---|---|---|---|
| 2026 | Diffusion Reconstruction-Based Data Likelihood Estimation for Core-Set Selection | AAAI 2026 | Data Distillation / Diffusion |
| 2026 | FedMCA: Cross-Attention-Based Personalized Multimodal Federated Learning | ICMSSP 2026 | Multimodal Medical FL |
| 2025 | Robustness Feature Adapter for Efficient Adversarial Training | ECAI 2025 | Efficient Adv. Training / PEFT |
| 2025 | Influence-Guided Diffusion for Dataset Distillation | ICLR 2025 | Generative Dataset Distillation |
| 2025 | StableGuard: Towards Unified Copyright Protection and Tamper Localisation | NeurIPS 2025 | AI Safety & Watermarking |
| 2024 | Federated Learning with Hybrid Knowledge Distillations on Long-Tailed Data | ECAI 2024 | Long-Tailed Federated Learning |
| 2023 | Boosting Accuracy and Robustness via Adaptive Adversarial Distillation | CVPR 2023 | Adversarial Knowledge Distillation |
| 2022 | Individual Property Inference Over Collaborative Learning | IEEE ICME 2022 | Privacy Leakage in FL |
| 2021 | DAIR: A Query-Efficient Decision-Based Attack On Image Retrieval | ACM SIGIR 2021 | Black-Box CBIR Attack |
| 2021 | Adversarial Defence By Diversified Simultaneous Training Of Deep Ensembles | AAAI 2021 | Deep Ensemble Robustness |
使用说明: 点击上方
<details>折叠区域标题即可展开对应语言版本。在支持 HTML 渲染的 Markdown 阅读器(如 GitHub、VS Code 预览、Typora)中效果最佳。Usage: Click the
<details>section headers above to expand the language version you need. Works best in Markdown renderers with HTML support (GitHub, VS Code preview, Typora, etc.).