PCEP: Few-Shot Model-Based Source Camera Identification

被引:3
|
作者
Wang, Bo [1 ]
Yu, Fei [1 ]
Ma, Yanyan [1 ]
Zhao, Haining [1 ]
Hou, Jiayao [1 ]
Zheng, Weiming [1 ]
机构
[1] Dalian Univ Technol, Sch Informat & Commun Engn, Dalian 116024, Peoples R China
基金
中国国家自然科学基金;
关键词
source camera identification; few-shot; prototype construction; ensemble projection; INVARIANT TEXTURE CLASSIFICATION; VIRTUAL SAMPLE GENERATION;
D O I
10.3390/math11040803
中图分类号
O1 [数学];
学科分类号
0701 ; 070101 ;
摘要
Source camera identification is an important branch in the field of digital forensics. Most existing works are based on the assumption that the number of training samples is sufficient. However, in practice, it is unrealistic to obtain a large amount of labeled samples. Therefore, in order to solve the problem of low accuracy for existing methods in a few-shot scenario, we propose a novel identification method called prototype construction with ensemble projection (PCEP). In this work, we extract a variety of features from few-shot datasets to obtain rich prior information. Then, we introduce semi-supervised learning to complete the construction of prototype sets. Subsequently, we use the prototype sets to retrain SVM classifiers, and take the posterior probability of each image sample belonging to each class as the final projection vector. Finally, we obtain classification results through ensemble learning voting. The PCEP method combines feature extraction, feature projection, classifier training and ensemble learning into a unified framework, which makes full use of image information of few-shot datasets. We conduct comprehensive experiments on multiple benchmark databases (i.e., Dresden, VISION and SOCRatES), and empirically show that our method achieves satisfactory performance and outperforms many recent methods in a few-shot scenario.
引用
收藏
页数:16
相关论文
共 50 条
  • [1] Virtual sample generation for few-shot source camera identification
    Wang, Bo
    Wu, Shiqi
    Wei, Fei
    Wang, Yue
    Hou, Jiayao
    Sui, Xue
    JOURNAL OF INFORMATION SECURITY AND APPLICATIONS, 2022, 66
  • [2] Few-Shot Model-Based Adaptation in Noisy Conditions
    Arndt, Karol
    Ghadirzadeh, Ali
    Hazara, Murtaza
    Kyrki, Ville
    IEEE ROBOTICS AND AUTOMATION LETTERS, 2021, 6 (02): : 4193 - 4200
  • [3] MDM-CPS: A few-shot sample approach for source camera identification
    Wang, Bo
    Hou, Jiayao
    Wei, Fei
    Yu, Fei
    Zheng, Weiming
    EXPERT SYSTEMS WITH APPLICATIONS, 2023, 229
  • [4] Voltage Sag Source Identification Based on Few-Shot Learning
    Sun, Haotian
    Yi, Hao
    Yang, Guangyu
    Zhuo, Fang
    Hu, Anping
    IEEE ACCESS, 2019, 7 : 164398 - 164406
  • [5] Multi-DS Strategy for Source Camera Identification in Few-Shot Sample Data Sets
    Wang, Bo
    Hou, Jiayao
    Ma, Yanyan
    Wang, Fei
    Wei, Fei
    SECURITY AND COMMUNICATION NETWORKS, 2022, 2022
  • [6] Multi-DS Strategy for Source Camera Identification in Few-Shot Sample Data Sets
    Wang, Bo
    Hou, Jiayao
    Ma, Yanyan
    Wang, Fei
    Wei, Fei
    Security and Communication Networks, 2022, 2022
  • [7] FEW-SHOT MUSICAL SOURCE SEPARATION
    Wang, Yu
    Stoller, Daniel
    Bittner, Rachel M.
    Bello, Juan Pablo
    2022 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH AND SIGNAL PROCESSING (ICASSP), 2022, : 121 - 125
  • [8] Few-Shot System Identification for Reinforcement Learning
    Farid, Karim
    Sakr, Nourhan
    2021 6TH ASIA-PACIFIC CONFERENCE ON INTELLIGENT ROBOT SYSTEMS (ACIRS), 2021, : 76 - 82
  • [9] CCFE: A Few-Shot Learning Model for Earthquake Detection and Phase Identification
    Zhao, Peng
    Yongqian, Shao
    Shian, Xia
    IEEE ACCESS, 2022, 10 : 124629 - 124636
  • [10] Few-shot Incremental Identification of Specific Emitter Based on Ancillary Prototypes
    Shi, Wenqiang
    Lei, Yingke
    Jin, Hu
    Teng, Fei
    Liu, Changming
    2024 IEEE/CIC INTERNATIONAL CONFERENCE ON COMMUNICATIONS IN CHINA, ICCC, 2024,