Virtual Mixup Training for Unsupervised Domain Adaptation

Abstract

We study the problem of unsupervised domain adaptation which aims to adapt models trained on a labeled source domain to a completely unlabeled target domain. Recently, the cluster assumption has been applied to unsupervised domain adaptation and achieved strong performance. One critical factor in successful training of the cluster assumption is to impose the locally-Lipschitz constraint to the model. Existing methods only impose the locally-Lipschitz constraint around the training points while miss the other areas, such as the points in-between training data. In this paper, we address this issue by encouraging the model to behave linearly in-between training points. We propose a new regularization method called Virtual Mixup Training (VMT), which is able to incorporate the locally-Lipschitz constraint to the areas in-between training data. Unlike the traditional mixup model, our method constructs the combination samples without using the label information, allowing it to apply to unsupervised domain adaptation. The proposed method is generic and can be combined with most existing models such as the recent state-of-the-art model called VADA. Extensive experiments demonstrate that VMT significantly improves the performance of VADA on six domain adaptation benchmark datasets. For the challenging task of adapting MNIST to SVHN, VMT can improve the accuracy of VADA by over 30\%. Code is available at \url{https://github.com/xudonmao/VMT}.

References in corpus (13)

• Learning Transferable Features with Deep Adaptation Networks
• MixMatch: A Holistic Approach to Semi-Supervised Learning
• Domain Separation Networks
• Unsupervised Image-to-Image Translation Networks
• Domain Adaptation: Learning Bounds and Algorithms
• Manifold Mixup: Better Representations by Interpolating Hidden States
• Asymmetric Tri-training for Unsupervised Domain Adaptation
• A DIRT-T Approach to Unsupervised Domain Adaptation
• Unsupervised Cross-Domain Image Generation
• Formal Guarantees on the Robustness of a Classifier against Adversarial Manipulation
• Co-regularized Alignment for Unsupervised Domain Adaptation
• Sample-to-Sample Correspondence for Unsupervised Domain Adaptation
• Unpaired Multi-Domain Image Generation via Regularized Conditional GANs

Cited by in corpus (13)

• Context-Aware Mixup for Domain Adaptive Semantic Segmentation
• Self-Supervised Learning for Domain Adaptation on Point-Clouds
• FixBi: Bridging Domain Spaces for Unsupervised Domain Adaptation
• Learning Domain Invariant Representations by Joint Wasserstein Distance Minimization
• Learning Smooth Representation for Unsupervised Domain Adaptation
• A Balanced and Uncertainty-aware Approach for Partial Domain Adaptation
• Integrating Categorical Semantics into Unsupervised Domain Translation
• AFAN: Augmented Feature Alignment Network for Cross-Domain Object Detection
• Semixup: In- and Out-of-Manifold Regularization for Deep Semi-Supervised Knee Osteoarthritis Severity Grading from Plain Radiographs
• SketchTransfer: A Challenging New Task for Exploring Detail-Invariance and the Abstractions Learned by Deep Networks
• MetaMix: Improved Meta-Learning with Interpolation-based Consistency Regularization
• Adversarial Mixing Policy for Relaxing Locally Linear Constraints in Mixup
• Unsupervised Adaptive Semantic Segmentation with Local Lipschitz Constraint