Research map / Label-efficient and robust learning / Domain adaptation and generalization

Revisiting Source-Free Domain Adaptation: a New Perspective via Uncertainty Control

ICLR 2025 · Label-efficient and robust learning › Domain adaptation and generalization · approach: Source-free domain adaptation

Read the paperCheck your idea against 108,967 papers

AI-extracted summary. Written by AI from the paper's abstract only. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Revisits source-free domain adaptation from an uncertainty-control perspective, targeting the uncertainty introduced during source-to-target transfer that SOTA SFDA methods overlook.

Problem

Source-free domain adaptation adapts a pre-trained source model to unlabeled target data without source data; current SOTA methods leverage weak supervision from the source model but often ignore the uncertainty arising during transfer.

Approach

Frames SFDA through explicit control of uncertainty during the transfer process, rather than only extracting reliable pseudo-supervision from the source model. Details beyond the abstract are not available.

Closest papers by meaning

• Understanding and Improving Source-free Domain Adaptation from a Theoretical Perspective (CVPR 2024 · similarity 0.93)
• Revisiting Source-Free Domain Adaptation: Insights into Representativeness, Generalization, and Variety (CVPR 2025 · similarity 0.92)
• Diffusion-Driven Progressive Target Manipulation for Source-Free Domain Adaptation (NeurIPS 2025 · similarity 0.91)
• Proxy Denoising for Source-Free Domain Adaptation (ICLR 2025 · similarity 0.90)
• Revisiting the Domain Shift and Sample Uncertainty in Multi-source Active Domain Transfer (CVPR 2024 · similarity 0.90)
• Dynamic Target Distribution Estimation for Source-Free Open-Set Domain Adaptation (AAAI 2025 · similarity 0.89)