Recall and Refine: A Simple but Effective Source-free Open-set Domain Adaptation Framework

2024-11-19

arXiv.org

AI Summary

We address source-free open-set domain adaptation (SF-OSDA), where source data are unavailable and the target domain contains unknown classes. Existing methods rely on entropy-based thresholds for unknown-class detection, failing to explicitly model discriminative features of unknown classes. We propose a two-stage paradigm: (1) boundary expansion via feature synthesis to explicitly model target-private unknown classes; and (2) joint optimization of domain alignment and unknown-class discrimination. Our method is framework-agnostic—compatible with SHOT, AaD, and others—and integrates boundary-augmented training with target-domain self-supervised fine-tuning. Evaluated on three standard benchmarks, it significantly outperforms state-of-the-art SF-OSDA and OSDA approaches, simultaneously improving both known-class accuracy and unknown-class identification. To our knowledge, this is the first method enabling robust, source-free discriminative modeling of unknown classes.

Technology Category

• Machine Learning: Transfer, Domain Adaptation, Multi-Task Learning
• Search and Optimization: Mixed Discrete/Continuous Search
• Computer Vision: Adversarial Attacks & Robustness

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
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs

Abstract

Open-set Domain Adaptation (OSDA) aims to adapt a model from a labeled source domain to an unlabeled target domain, where novel classes - also referred to as target-private unknown classes - are present. Source-free Open-set Domain Adaptation (SF-OSDA) methods address OSDA without accessing labeled source data, making them particularly relevant under privacy constraints. However, SF-OSDA presents significant challenges due to distribution shifts and the introduction of novel classes. Existing SF-OSDA methods typically rely on thresholding the prediction entropy of a sample to identify it as either a known or unknown class but fail to explicitly learn discriminative features for the target-private unknown classes. We propose Recall and Refine (RRDA), a novel SF-OSDA framework designed to address these limitations by explicitly learning features for target-private unknown classes. RRDA employs a two-step process. First, we enhance the model's capacity to recognize unknown classes by training a target classifier with an additional decision boundary, guided by synthetic samples generated from target domain features. This enables the classifier to effectively separate known and unknown classes. In the second step, we adapt the entire model to the target domain, addressing both domain shifts and improving generalization to unknown classes. Any off-the-shelf source-free domain adaptation method (e.g., SHOT, AaD) can be seamlessly integrated into our framework at this stage. Extensive experiments on three benchmark datasets demonstrate that RRDA significantly outperforms existing SF-OSDA and OSDA methods.

Problem

• Adapting model to target domain with novel classes
• Learning discriminative features for unknown classes
• Addressing domain shifts without source data access

Innovation

• Two-stage SF-OSDA framework for unknown classes
• Synthetic samples guide decision boundary training
• Seamless integration with existing domain adaptation methods