A Prototype-Oriented Framework for Unsupervised Domain Adaptation

2021/10/22 by Korawat Tanwisuth, Tanwisuth, Korawat, Xinjie Fan +11 · 5 citations
Computer Science · Mathematics · #Computer Vision and Pattern Recognition (cs.CV) #Domain Adaptation and Few-Shot Learning #FOS: Computer and information sciences #Machine Learning (cs.LG) #Machine Learning (stat.ML) #Multimodal Machine Learning Applications #cs.CV #cs.LG #stat.ML

paper · pdf · doi:10.48550/arxiv.2110.12024

NeurIPS 2021

arxiv created 2021/10/22 · openalex publication_date 2021/10/22 · arxiv updated 2021/10/26 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

Existing methods for unsupervised domain adaptation often rely on minimizing some statistical distance between the source and target samples in the latent space. To avoid the sampling variability, class imbalance, and data-privacy concerns that often plague these methods, we instead provide a memory and computation-efficient probabilistic framework to extract class prototypes and align the target features with them. We demonstrate the general applicability of our method on a wide range of scenarios, including single-source, multi-source, class-imbalance, and source-private domain adaptation. Requiring no additional model parameters and having a moderate increase in computation over the source model alone, the proposed method achieves competitive performance with state-of-the-art methods.

Citations

• Deep Domain Confusion: Maximizing for Domain Invariance
• Generative Adversarial Networks
• Adjusting the Outputs of a Classifier to New aPrioriProbabilities: A Simple Procedure
• Learning Transferable Features with Deep Adaptation Networks
• Unsupervised Domain Adaptation with Residual Transfer Networks
• Prototypical Networks for Few-shot Learning
• Maximum Classifier Discrepancy for Unsupervised Domain Adaptation
• MONK -- Outlier-Robust Mean Embedding Estimation by Median-of-Means
• Conditional Adversarial Domain Adaptation
• On Causal and Anticausal Learning
• Temporal Ensembling for Semi-Supervised Learning
• ADVENT: Adversarial Entropy Minimization for Domain Adaptation in Semantic Segmentation
• Entropy Minimization vs. Diversity Maximization for Domain Adaptation
• Do We Really Need to Access the Source Data? Source Hypothesis Transfer for Unsupervised Domain Adaptation
• Exploring Simple Siamese Representation Learning

Cited by

• A Tale of Two Experts: Cooperative Learning for Source-Free Unsupervised Domain Adaptation
• Frustratingly Easy Feature Reconstruction for Out-of-Distribution Detection
• Doodle Your Keypoints: Sketch-Based Few-Shot Keypoint Detection
• A Comprehensive Survey on Test-Time Adaptation Under Distribution Shifts
• Context Aware Grounded Teacher for Source Free Object Detection

Related

• Feed-Forward Source-Free Domain Adaptation via Class Prototypes
• Prototype-Guided Continual Adaptation for Class-Incremental Unsupervised Domain Adaptation
• Multi-Source Unsupervised Domain Adaptation with Prototype Aggregation
• Source-free Domain Adaptation via Avatar Prototype Generation and Adaptation
• Domain Impression: A Source Data Free Domain Adaptation Method