1S-DAug: One-Shot Data Augmentation for Robust Few-Shot Generalization

Yunwei Bai, Ying Kiat Tan, Yao Shu, Tsuhan Chen · 2026

Read 1S-DAug: One-Shot Data Augmentation for Robust Few-Shot Generalization by Yunwei Bai, Ying Kiat Tan, Yao Shu, Tsuhan Chen (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2602-00114.txt.

Description

Yunwei Bai and colleagues ask whether test time augmentation from a single image can help models generalize. Most augmentation happens in training, while test time use is less studied, though it avoids extra weights or tuning. Their operator, 1S-DAug, joins geometric shifts with controlled noise and image guided denoising. Tests on four image classification sets across several models lift accuracy by up to a fifth in relative terms, with no training or weight access needed.

Catalog record

CREATORYunwei Bai, Ying Kiat Tan, Yao Shu, Tsuhan Chen
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE72926
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2602.00114

Browse the full library · About the Tower of Babel