Virtual sample generation for few-shot source camera identification

doi:10.1016/j.jisa.2022.103153

Virtual sample generation for few-shot source camera identification

2022-05-01

6

PRE

AI

DOI：10.1016/j.jisa.2022.103153

Original

Original request for help

Save

Abstract

En 中文

The Source Camera Identification (SCI) has achieved remarkable success. However, existing approachesrequire sufficiently large training sets for high performance on accuracy and robustness. For maintaining highperformance given small training sets, we propose a semi-supervised, Mega-Trent-Diffusion (MTD) method togenerate virtual samples, such that the training sets can be expanded and unlabeled samples can be fullyutilized as well. The stability of our method is improved using ensemble learning. Our theoretical analysis andexperiments corroborate the effectiveness of our method beyond others when few-shot is given

Keywords:

Source Camera Identification (SCI)

Few-shot

Virtual sample

Ensemble learning

Semi-supervised

Virtual sample generation for few-shot source camera identification

Abstract

Journal

Organization