Associative Domain Adaptation

Abstract

We propose associative domain adaptation, a novel technique for end-to-end domain adaptation with neural networks, the task of inferring class labels for an unlabeled target domain based on the statistical properties of a labeled source domain. Our training scheme follows the paradigm that in order to effectively derive class labels for the target domain, a network should produce statistically domain invariant embeddings, while minimizing the classification error on the labeled source domain. We accomplish this by reinforcing associations between source and target data directly in embedding space. Our method can easily be added to any existing classification network with no structural and almost no computational overhead. We demonstrate the effectiveness of our approach on various benchmarks and achieve state-of-the-art results across the board with a generic convolutional neural network architecture not specifically tuned to the respective tasks. Finally, we show that the proposed association loss produces embeddings that are more effective for domain adaptation compared to methods employing maximum mean discrepancy as a similarity measure in embedding space.

In IEEE International Conference on Computer Vision (ICCV), 2017

References in corpus (7)

• DeepPose: Human Pose Estimation via Deep Neural Networks
• Going Deeper with Convolutions
• Beyond Sharing Weights for Deep Domain Adaptation
• Deep CORAL: Correlation Alignment for Deep Domain Adaptation
• Return of Frustratingly Easy Domain Adaptation
• Unsupervised Pixel-Level Domain Adaptation with Generative Adversarial Networks
• Learning by Association - A versatile semi-supervised training method for neural networks

Cited by in corpus (14)

• Deep Visual Domain Adaptation: A Survey
• Maximum Classifier Discrepancy for Unsupervised Domain Adaptation
• Self-ensembling for visual domain adaptation
• Deep Cocktail Network: Multi-source Unsupervised Domain Adaptation with Category Shift
• Minimal-Entropy Correlation Alignment for Unsupervised Deep Domain Adaptation
• Text2Shape: Generating Shapes from Natural Language by Learning Joint Embeddings
• Deep Adversarial Attention Alignment for Unsupervised Domain Adaptation: the Benefit of Target Expectation Maximization
• From source to target and back: symmetric bi-directional adaptive GAN
• Unsupervised Domain Adaptation for Spatio-Temporal Action Localization
• Zero-Shot Deep Domain Adaptation
• Domain Agnostic Real-Valued Specificity Prediction
• Unsupervised Feature Selection via Multi-step Markov Transition Probability
• Learning Condensed and Aligned Features for Unsupervised Domain Adaptation Using Label Propagation
• Dynamic Adaptation on Non-Stationary Visual Domains