6 papers
Dynamic Epsilon Scheduling: A Multi-Factor Adaptive Perturbation Budget for Adversarial Training
Alan Mitkiy, James Smith, Myungseo wong +3
Adversarial training is among the most effective strategies for defending deep neural networks against adversarial examples. A key limitation of existing adversarial training appro…
GAMA: Geometry-Aware Manifold Alignment via Structured Adversarial Perturbations for Robust Domain Adaptation
Hana Satou, F Monkey
Domain adaptation remains a challenge when there is significant manifold discrepancy between source and target domains. Although recent methods leverage manifold-aware adversarial…
On the Mechanisms of Adversarial Data Augmentation for Robust and Adaptive Transfer Learning
Hana Satou, Alan Mitkiy
Transfer learning across domains with distribution shift remains a fundamental challenge in building robust and adaptable machine learning systems. While adversarial perturbations…
Disentangled Geometric Alignment with Adaptive Contrastive Perturbation for Reliable Domain Transfer
Emma Collins, Myungseo wong, Kim Yun +2
Despite progress in geometry-aware domain adaptation, current methods such as GAMA still suffer from two unresolved issues: (1) insufficient disentanglement of task-relevant and ta…
Geometrically Regularized Transfer Learning with On-Manifold and Off-Manifold Perturbation
Hana Satou, Alan Mitkiy, Emma Collins +1
Transfer learning under domain shift remains a fundamental challenge due to the divergence between source and target data manifolds. In this paper, we propose MAADA (Manifold-Aware…
Fusing Physics-Driven Strategies and Cross-Modal Adversarial Learning: Toward Multi-Domain Applications
Hana Satou, Alan Mitkiy
The convergence of cross-modal adversarial learning and physics-driven methods represents a cutting-edge direction for tackling challenges in complex multi-modal tasks and scientif…