半監督噪聲適應：從噪聲域遷移知識

摘要

遷移學習旨在通過從源域遷移知識來促進目標域的學習。源域通常包含語義上有意義的樣本（例如圖像），以實現有效的知識遷移。然而，近期一項研究觀察到，由簡單分佈（例如高斯分佈）構成的噪聲域，可在半監督設定下作為替代源域——該設定中僅有少量目標樣本被標記，其餘多數保持未標記狀態。基於此驚人發現，我們提出一個新問題，稱為「半監督噪聲適應」（Semi-Supervised Noise Adaptation, SSNA），旨在利用合成的噪聲域提升目標域的泛化能力。為解決該問題，我們首先建立一個刻畫噪聲域對泛化影響的泛化界限，並在此基礎上提出噪聲適應框架（Noise Adaptation Framework, NAF）。大量實驗表明，NAF能有效利用噪聲域以收緊目標域的泛化界限，從而帶來性能提升。相關代碼已公開於 https://github.com/AIResearch-Group/SSNA。

English

Transfer learning aims to facilitate the learning of a target domain by transferring knowledge from a source domain. The source domain typically contains semantically meaningful samples (*e.g.*, images) to facilitate effective knowledge transfer. However, a recent study observes that the noise domain constructed from simple distributions (*e.g.*, Gaussian distributions) can serve as a surrogate source domain in the semi-supervised setting, where only a small proportion of target samples are labeled while most remain unlabeled. Based on this surprising observation, we formulate a novel problem termed *Semi-Supervised Noise Adaptation* (SSNA), which aims to leverage a synthetic noise domain to improve the generalization of the target domain. To address this problem, we first establish a generalization bound characterizing the effect of the noise domain on generalization, based on which we propose a Noise Adaptation Framework (NAF). Extensive experiments demonstrate that NAF effectively leverages the noise domain to tighten the generalization bound of the target domain, leading to improved performance. The codes are available at https://github.com/AIResearch-Group/SSNA.