GAMA: Geometry-Aware Manifold Alignment via Structured Adversarial Perturbations for Robust Domain Adaptation

2025/05/21 by Hana Satou, Satou, Hana, F Monkey +1
Computer Science · #Domain Adaptation and Few-Shot Learning #Human Pose and Action Recognition #Anomaly Detection Techniques and Applications

paper · pdf · doi:10.48550/arxiv.2505.15194

Abstract

Domain adaptation remains a challenge when there is significant manifold discrepancy between source and target domains. Although recent methods leverage manifold-aware adversarial perturbations to perform data augmentation, they often neglect precise manifold alignment and systematic exploration of structured perturbations. To address this, we propose GAMA (Geometry-Aware Manifold Alignment), a structured framework that achieves explicit manifold alignment via adversarial perturbation guided by geometric information. GAMA systematically employs tangent space exploration and manifold-constrained adversarial optimization, simultaneously enhancing semantic consistency, robustness to off-manifold deviations, and cross-domain alignment. Theoretical analysis shows that GAMA tightens the generalization bound via structured regularization and explicit alignment. Empirical results on DomainNet, VisDA, and Office-Home demonstrate that GAMA consistently outperforms existing adversarial and adaptation methods in both unsupervised and few-shot settings, exhibiting superior robustness, generalization, and manifold alignment capability.

Citations

• Cross-Modality Attack Boosted by Gradient-Evolutionary Multiform Optimization
• Adversarial Learning for Neural PDE Solvers with Sparse Data
• Beyond Augmentation: Empowering Model Robustness under Extreme Capture Environments
• Cross-Task Attack: A Self-Supervision Generative Framework Based on Attention Shift
• Beyond Dropout: Robust Convolutional Neural Networks Based on Local Feature Masking
• Exploring Color Invariance through Image-Level Ensemble Learning
• Cross-Modality Perturbation Synergy Attack for Person Re-identification
• Eliminate Deviation with Deviation for Data Augmentation and a General Multi-modal Data Learning Method
• Mutual Mean-Teaching: Pseudo Label Refinery for Unsupervised Domain Adaptation on Person Re-identification
• Adversarial Training for Free!
• Bridging Theory and Algorithm for Domain Adaptation
• Theoretically Principled Trade-off between Robustness and Accuracy
• Moment Matching for Multi-Source Domain Adaptation
• Disentangling Adversarial Robustness and Generalization
• VisDA: The Visual Domain Adaptation Challenge
• Deep Hashing Network for Unsupervised Domain Adaptation
• Towards Deep Learning Models Resistant to Adversarial Attacks
• Virtual Adversarial Training: A Regularization Method for Supervised and Semi-Supervised Learning
• Explaining and Harnessing Adversarial Examples

Related

• Disentangled Geometric Alignment with Adaptive Contrastive Perturbation for Reliable Domain Transfer
• Geometrically Regularized Transfer Learning with On-Manifold and Off-Manifold Perturbation
• Generating and Aligning from Data Geometries with Generative Adversarial Networks
• Unsupervised Domain Adaptation via Discriminative Manifold Embedding and Alignment
• Filtered Manifold Alignment