Skip to main content
SpringerLink
Log in

Detecting double JPEG compression and its related anti-forensic operations with CNN

  • Published:
Multimedia Tools and Applications Aims and scope Submit manuscript

Abstract

Detecting double JPEG compression is important to forensic experts in identifying the originality and authenticity of images. However, there are some anti-forensic techniques which can evade existing double compression detectors. It is desirable to design a unified approach to address the issues of JPEG forensics and counter-anti-forensics simultaneously, but existing hand-crafted feature based methods and deep learning based methods may fail to satisfy the requirement. In this paper, we present a data-driven approach by using a convolutional neural network (CNN) which takes input from both raw JPEG DCT coefficients and decompressed image pixels. Expert knowledge about JPEG characteristics is incorporated in the CNN design by exploring the intricate relations both within and among DCT subbands and by looking for spatial artifacts both within and among JPEG grids. The CNN is capable of learning deep representations from training data and thus can effectively detect double JPEG compression and its related anti-forensic operations together. The end-to-end CNN that takes into account the information from both DCT domain and spatial domain, shows outstanding performance when compared to prior arts in the experiments. It shows a promising way to address counter-anti-forensic issues without designing specific features for each anti-forensic operation.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6
Fig. 7
Fig. 8
Fig. 9

Notes

  1. We have verified that the same CNN structure with the input size of 256 × 256 works better than the input size of 64 × 64 (by using the center patch of an image). We also compare the performance with and without dropout, and adopt the implementation without dropout since it converges more fast and has similar performance as the implementation with dropout.

References

  1. Abadi M, Agarwal A, Barham P, Brevdo E, Chen Z, Citro C, Corrado GS, Davis A, Dean J, Devin M et al (2016) Tensorflow: large-scale machine learning on heterogeneous distributed systems. arXiv:160304467

  2. Amerini I, Becarelli R, Caldelli R, Del Mastio A (2014) Splicing forgeries localization through the use of first digit features. In: Proc. IEEE Int. workshop inf. forensics secur., pp 143–148

  3. Amerini I, Uricchio T, Ballan L, Roberto Caldelli I et al (2017) Localization of JPEG double compression through multi-domain convolutional neural networks. In: Proc. the IEEE conference on computer vision and pattern recognition workshops, pp 53–59

  4. Anthony HT, Li S (2015) Handbook of digital forensics of multimedia data and devices. Wiley

  5. Barni M, Bondi L, Bonettini N, Bestagini P, Costanzo A, Maggini M, Tondi B, Tubaro S (2017) Aligned and non-aligned double JPEG detection using convolutional neural networks. J Vis Commun Image Represent 49:153–163

    Article  Google Scholar 

  6. Barni M, Nowroozi E, Tondi B (2017) Higher-order, adversary-aware, double JPEG-detection via selected training on attacked samples. In: Proc. 25th European Signal Processing Conference (EUSIPCO), pp 281–285

  7. Bas P, Filler T, Pevnỳ T (2011) Break our steganographic system: The ins and outs of organizing boss. In: Proc. inf. hiding, pp 59–70

  8. Bianchi T, Piva A (2012) Image forgery localization via block-grained analysis of JPEG artifacts. IEEE Trans Inf Forens Secur 7(3):1003–1017

    Article  Google Scholar 

  9. Boureau YL, Bach F, LeCun Y, Ponce J (2010) Learning mid-level features for recognition. In: Proc. IEEE Conference on computer vision and pattern recognition (CVPR), pp 2559–2566

  10. Chu X, Stamm M, Chen Y, Liu K (2015) On antiforensic concealability with rate-distortion tradeoff. IEEE Trans Image Process 24(3):1087–1100

    Article  MathSciNet  Google Scholar 

  11. Clevert D, Unterthiner T, Hochreiter S (2015) Fast and accurate deep network learning by exponential linear units(elus). arXiv:151107289

  12. Fan W, Wang K, Cayre F, Xiong Z (2013) JPEG anti-forensics using non-parametric DCT quantization noise estimation and natural image statistics. In: Proc. the 1st ACM workshop on inf. hiding and multimedia secur., pp 117–122

  13. Fan W, Wang K, Cayre F, Xiong Z (2013) A variational approach to JPEG anti-forensics. In: Proc. IEEE international conference on acoustics, speech and signal processing (ICASSP), pp 3058–3062

  14. Fan W, Wang K, Cayre F, Xiong Z (2014) JPEG anti-forensics with improved tradeoff between forensic undetectability and image quality. IEEE Trans Inf Forens Secur 9(8):1211–1226

    Article  Google Scholar 

  15. Fridrich J, et al. (2008) Detection of double-compression in JPEG images for applications in steganography. IEEE Trans Inf Forens Secur 3(2):247–258

    Article  Google Scholar 

  16. Fu D, Shi YQ, Su W (2007) A generalized Benford’s law for JPEG coefficients and its applications in image forensics. In: Proc. SPIE Electronic imaging, security and watermarking of multimedia contents IX, vol 6505, pp 1L1–1L11

  17. Glorot X, Bengio Y (2010) Understanding the difficulty of training deep feedforward neural networks. J Mach Learn Res 9:249–256

    Google Scholar 

  18. Glorot X, Bordes A, Bengio Y (2012) Deep sparse rectifier neural networks. In: Proc. International conference on artificial intelligence and statistics, pp 315–323

  19. He K, Zhang X, Ren S, Sun J (2016) Deep residual learning for image recognition. In: Proc. IEEE Conference on computer vision and pattern recognition (CVPR), pp 770–778

  20. Ioffe S, Szegedy C (2015) Batch normalization: accelerating deep network training by reducing internal covariate shift. In: Proc. International conference on machine learning, pp 448–456

  21. Kalman BL, Kwasny SC (1992) Why tanh: choosing a sigmoidal function. In: Proc. International joint conference on neural networks (IJCNN), vol 4, pp 578–581

  22. Krizhevsky A, Sutskever I, Hinton GE (2012) Imagenet classification with deep convolutional neural networks. In: Proc. International conference on neural information processing systems, pp 1097–1105

  23. Lai S, Böhme R (2011) Countering counter-forensics: the case of JPEG compression. In: Proc. inf. hiding, pp 285–298

  24. LeCun Y, Bottou L, Bengio Y, Haffner P (1998) Gradient-based learning applied to document recognition. In: Proc. IEEE, vol 86, pp 2278–2324

    Article  Google Scholar 

  25. Li B, Shi YQ, Huang J (2008) Detecting doubly compressed JPEG images by using mode based first digit features. In: Proc. IEEE workshop on multimedia signal processing, pp 730–735

  26. Li B, Luo H, Zhang H, Tan S, Ji Z (2017) A multi-branch convolutional neural network for detecting double JPEG compression. arXiv:171005477

  27. Li B, Li Z, Zhou S, Tan S, Zhang X (2018) New steganalytic features for spatial image steganography based on derivative filters and threshold LBP operator. IEEE Trans Inf Forens Secur 13(5):1242–1257

    Article  Google Scholar 

  28. Li H, Luo W, Huang J (2012) Countering anti-JPEG compression forensics. In: Proc. 19th IEEE international conference on image processing (ICIP), pp 241–244

  29. Lukás J, Fridrich J (2003) Estimation of primary quantization matrix in double compressed JPEG images. In: Proc. digital forensic research workshop. Cleveland, pp 5–8

  30. Pasquini C, Boato G, Perez-Gonzalez F (2017) Statistical detection of JPEG traces in digital images in uncompressed formats. IEEE Trans Inf Forens Secur 12 (12):2890–2905

    Article  Google Scholar 

  31. Piva A (2013) An overview on image forensics. ISRN Signal Process 2013:1–22

    Article  Google Scholar 

  32. Popescu AC, Farid H (2004) Statistical tools for digital forensics. Springer, Berlin

    Book  Google Scholar 

  33. Simonyan K, Zisserman A (2014) Very deep convolutional networks for large-scale image recognition. arXiv:14091556

  34. Singh G, Singh K (2017) Improved JPEG anti-forensics with better image visual quality and forensic undetectability. Forensic Science International

  35. Stamm MC, Liu KR (2011) Anti-forensics of digital image compression. IEEE Trans Inf Forens Secur 6(3):1050–1065

    Article  Google Scholar 

  36. Stamm MC, Tjoa SK, Lin WS, Liu KR (2010) Anti-forensics of JPEG compression. In: Proc. IEEE International conference on acoustics speech and signal processing (ICASSP), pp 1694–1697

  37. Stamm MC, Tjoa SK, Lin WS, Liu KR (2010) Undetectable image tampering through JPEG compression anti-forensics. In: Proc. 17th IEEE International conference on image processing (ICIP), pp 2109–2112

  38. Taimori A, Razzazi F, Behrad A, Ahmadi A, Babaie-Zadeh M (2016) Quantization-unaware double JPEG compression detection. J Math Imag Vis 54(3):269–286

    Article  MathSciNet  Google Scholar 

  39. Valenzise G, Nobile V, Tagliasacchi M, Tubaro S (2011) Countering JPEG anti-forensics. In: Proc. 18th IEEE international conference on image processing (ICIP), pp 1949–1952

  40. Valenzise G, Tagliasacchi M, Tubaro S (2011) The cost of JPEG compression anti-forensics. In: Proc. IEEE international conference on acoustics, speech and signal processing (ICASSP), pp 1884–1887

  41. Valenzise G, Tagliasacchi M, Tubaro S (2013) Revealing the traces of JPEG compression anti-forensics. IEEE Trans Inf Forens Secur 8(2):335–349

    Article  Google Scholar 

  42. Valenzise G, Tagliasacchi M, Tubaro S (2014) Detectability-quality trade-off in jpeg counter-forensics. In: IEEE International conference on image processing, pp 5337–5341

  43. Verma V, Agarwal N, Khanna N (2017) DCT-domain deep convolutional neural networks for multiple JPEG compression classification. arXiv:171202313

  44. Wallace GK (1991) The JPEG still picture compression standard. Commun ACM 34(4):30–44

    Article  Google Scholar 

  45. Wang Q, Zhang R (2016) Double JPEG compression forensics based on a convolutional neural network. EURASIP J Inf Secur 11(1):23

    Article  Google Scholar 

  46. Yu J, Zhan Y, Yang J, Kang X (2016) A multi-purpose image counter-anti-forensic method using convolutional neural networks. In: Proc. International workshop on digital watermarking, pp 3–15

    Chapter  Google Scholar 

Download references

Acknowledgements

This work was supported in part by NSFC (Grant 61872244, 61572329, 61772349, and U1636202) and in part by the Shenzhen R&D Program (Grant JCYJ20160328144421330).

Author information

Authors and Affiliations

  1. Guangdong Key Laboratory of Intelligent Information Processing and Shenzhen Key Laboratory of Media Security, Shenzhen University, Shenzhen, China

    Bin Li, Haoxin Zhang, Hu Luo & Shunquan Tan

Authors
  1. Bin Li
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Haoxin Zhang
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Hu Luo
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Shunquan Tan
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Bin Li.

Additional information

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Li, B., Zhang, H., Luo, H. et al. Detecting double JPEG compression and its related anti-forensic operations with CNN. Multimed Tools Appl 78, 8577–8601 (2019). https://doi.org/10.1007/s11042-018-7073-3

Download citation

  • Received:

  • Revised:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s11042-018-7073-3

Keywords

Search

Navigation