Noise Optimized Conditional Diffusion for Domain Adaptation

2025-05-12

AI Summary

To address inaccurate cross-domain statistical alignment in unsupervised domain adaptation (UDA) caused by the scarcity of high-confidence pseudo-labeled target samples, this paper proposes a task-coupled noise-optimized conditional diffusion model. Our method tackles the problem by: (1) introducing a novel class-aware noise initialization strategy—replacing standard Gaussian noise—to significantly enhance the generation quality of class-specific, high-confidence pseudo-labels during reverse sampling; and (2) jointly optimizing the diffusion classifier and the domain adaptation classifier, while incorporating cross-domain consistency regularization to ensure discriminative consistency throughout the noise evolution process. Extensive experiments across five benchmark datasets and 29 domain adaptation tasks demonstrate that our approach consistently outperforms 31 state-of-the-art methods, validating its effectiveness and robustness.

Technology Category

• Machine Learning: Transfer, Domain Adaptation, Multi-Task Learning
• Computer Vision: Diffusion Models for Vision
• Search and Optimization: Learning to Search

Application Category

• Search and Retrieval-Augmented AI: Vertical and domain-specific search
• User Modeling, Personalization and Recommendation: User modeling for targeted and personalized online advertising
• Economics, Online Markets and Human Computation: Trust and reliance of crowd workers and data experts on GenAI

Abstract

Pseudo-labeling is a cornerstone of Unsupervised Domain Adaptation (UDA), yet the scarcity of High-Confidence Pseudo-Labeled Target Domain Samples ( extbf{hcpl-tds}) often leads to inaccurate cross-domain statistical alignment, causing DA failures. To address this challenge, we propose extbf{N}oise extbf{O}ptimized extbf{C}onditional extbf{D}iffusion for extbf{D}omain extbf{A}daptation ( extbf{NOCDDA}), which seamlessly integrates the generative capabilities of conditional diffusion models with the decision-making requirements of DA to achieve task-coupled optimization for efficient adaptation. For robust cross-domain consistency, we modify the DA classifier to align with the conditional diffusion classifier within a unified optimization framework, enabling forward training on noise-varying cross-domain samples. Furthermore, we argue that the conventional ( mathcal{N}(mathbf{0}, mathbf{I}) ) initialization in diffusion models often generates class-confused hcpl-tds, compromising discriminative DA. To resolve this, we introduce a class-aware noise optimization strategy that refines sampling regions for reverse class-specific hcpl-tds generation, effectively enhancing cross-domain alignment. Extensive experiments across 5 benchmark datasets and 29 DA tasks demonstrate significant performance gains of extbf{NOCDDA} over 31 state-of-the-art methods, validating its robustness and effectiveness.

Problem

• Addresses scarcity of high-confidence pseudo-labels in domain adaptation
• Proposes noise-optimized conditional diffusion for task-coupled DA optimization
• Improves cross-domain alignment via class-aware noise optimization strategy

Innovation

• Integrates conditional diffusion models with DA
• Modifies DA classifier for cross-domain consistency
• Introduces class-aware noise optimization strategy