Program of Academic Course

Introduction to Information Hiding

Doctor of Technical Science,

Renowned Worker of Russian FederationHigh School,

IEEE Member ,

Professor of State TelecommunicationUniversity in St.Petersburg (Russian Federation) , Valery Korzhik

Email:

Helsinki University, Finland

2014

Contents

Theme 1. Introduction to the lecture course

Two main parts of information hiding (IH): steganography (SG) and digital watermarking (WM). Terminology and classification. SG versus cryptography.The main types of cover objects (CO) .Criteria of SG system efficiency.The main attacks on SG .Historical review of SG.

Examples of the SG- use by terrorists and criminals. Traditional and new results in the current course.

Similar courses which are commonly to give in other Universities.

Theme 2 .SG with embedding in the least significant bits of CO.

Embedding and detecting procedure. Selection channel. Detectability of SG based on embedding in the LSB (histogram and sample pair analysis attacks).

Theme 3 .Real SG systems and spread spectrum(SS)- based SG.

SG in Real World (Software market).

Examples of SG systems (Jsteg , F5, Outguess ctr.).Syndrom embedding. Embedding and extraction with SG-SS .Blind and informed decoding. Detecting of SS-based SG.

Theme 4 .SG for other cover objects.

Embedding into image-documents (SG-SC). The performance evaluation of different SG schemes.Coumuflage of embedding by scanner noices.Linguistic steganography (SG-L).Embedding based on absolute and relative synonymous. Rejection-sampling embedding. Internet SG.

Internet-based steganography.

Theme 5 . Perfectly secure and almost perfectly secure SG .

Relative entropy and Bchattacharyya Distance(BD) as criteria of SG system security .Ideal SG for Gaussian CM. Model-based SG(SG-M).Perturbed quantization steganography (PQS) and their security. Wet paper codes. Hugo algorithm.SG parameter optimization based on the use of NND estimator.

Demonstration of PQS algorithm.

Theme 6 . Steganography based on noisy channels.

Setting of SG-N model based on noisy channels. BSC and Gaussian channel. Embedding and extraction procedures. Security and reliability of such SG systems. Spread-time SG (STS) and their security and reliability. SG-N systems with blind decoder.

Theme 7. Blind steganalysis .

Classification of steganalytic methods .Support vector machine. Linear and non-linear versions. Choice of statistical functionals. SPAM characteristics.Calibration of images.

Theme 8 .The main applications of digital watermarking.

Broadcast monitoring , owner identification, proof of ownership,

transaction tracking ,content authentication ,copy control, device control. Fingerprinting. Tracing traitors.

Theme 9 . Watermark embedding and detection technique .

Embedding into transformed domains (DFT , DCT, DWT). Additive and multiplicative embedding .Examples of robust WM : “golographic” –based (WM-G) and maxima-based WM(WM-M).Multiple and null bit WM.

Theme 10. Embedding with informed encoders.

Code book for 0-bit WM. Improved spread spectrum system(ISS).Quantization projective modulation (QPM).Comparison of ISS and QPM efficiency.

Theme 11 .Sophisticated attacks on WM.

Filtering attack. Estimation of WM. Compression/decompression attack.

Transforms D/A and A/D.

Theme 12 .Breaking of WM synchronization.Collusion attacks.Resizing,cropping andimage rotation. The use ofFourier-Mellin Transform.Concatenated WM. Different types of collusion attacks on WM systems .The performance evaluation of WM systems under the conditions of collusion attacks.

The problem of tracing traitors The use of anticollusion- coded WM in order to provide robustness of WM systems. Superimposed codes in real space.

System attacks. Unauthorized WM embedding. Copy attack.Ambiguity attack.

Theme 13 . Content authentication based on WM systems.

Problem of authentication based on WM systems(WM-AU). Modular embedding and its efficiency .Distortion free embedding for images in BMP and JPEG formats. Selective authentication based on WM.

Robustness to JPEG compression.

Theme 14 .Watermarking of audio signals.

Model of audio CM. Human auditory system (HAS).Masking in time and in frequency domains. Modified LSB embedding. Phase modulation embedding. Echo hiding.Error free embedding algorithm. Reverberation hiding. Modified ISS-based embedding.

Theme 15 .Capacity of SG and WM systems.

Definition of SG and WM capacity .Capacity of SG with noisy channels. Binary and Gaussian models of WM systems.

Discussion on WM and SG capacity.

Remark .The required background for the listeners of this course is some basic mathematical knowledge including elements of probability theory and elements of digital communications.

References

1.M.Maes ,”Twin Peaks: ”The Histogram Attack to Fixed Depth Image Watermarking,”

LNCS,1525,pp.290-305, 1998.

2.M.J.Atallah et al ,”Natural Language Watermarking and Tamperproofing “,LNCS 2578 ,pp.196-212,2003.

3.I.Melcuk , “Dependency Syntax : Theory and Practice.” SONY Press, NY, 1988.

4.J Brassil, L.O’ Gorman, “ Watermarking Document Images with Bounding Box Expansion”, Proc. First Int. Workshop on IH, pp. 227-235,1996.

5.T. Handel, M. Sandford, “Hiding Data in the OSI Network Model”, Proc. First Int. Workshop on IH, pp.23-38,1996.

6.T.M. Cover, J.A. Thomas, “Elements of Information Theory”, John Wiley, 1991.

7. C. Cachin, “ An information – theoretic model for steganography”, LNCS, 1525, pp.306-318, 1998.

8.T. Kailath, “ The divergence and Bhattacharyya distance measures in signal selection”, IEEE Trans. Commun. Tech. 15, pp. 52-60, 1967.

9. V. Korzhik et al. “ On the use of Bhattacharyya distance as a measure of detectability of steganographic systems”, Information Hiding International Journal, N 1, 2008, pp.23-30.

10. J. Wang, P. Moulin, “ Steganalysis of block-structured stegotext”. In: Security, Steganography and Watermarking of Multimedia Contents, SPIE Proc. Vol. 5306, pp.477-488, 2004.

11. P. Salee , “ Model – based steganography “. LNCS, 2939, pp. 154-167, 2004.

12. J. Fridrich , et al “ Perturbed quantization steganography with wet paper codes “. In: Proc. Workshop on Multimedia and Security. 2004, pp, 4-15, 2004.

13. V. Korzhik et al, “ Investigation of Some Attacks on Perturbed Quantization Steganography” ( submitted).

14. V. Korzhik, et al, “ On the existence of perfect stegosystems”, LNCS, 3710, pp.30- 37,2005.

15. V. Korzhik , et al. “ Stegosystems based on noisy channels “, Proc. IX Spanish Meeting on Cryptology and Information Security, 2006.

16. J. Fridrich, “ Feature – based steganalysis for JPEG images and its applications for future design of steganographic schemes “. LNCS, 3200, pp. 67-81,2005.

17. S. Lyu , H. Farid, “ Detecting Hidden Messages Using Higher – Order Statistics and Support Vector Machines “, In: Proc. 5th Int. Workshop on IH, 2002.

18. C. Burges , “ A Tutorial on Support Vector Machines for Pattern Recognition “, Data Mining and Knowledge Discovery, 2, pp. 121-167, 1998.

19. I, Cox, et al “ Digital Watermarking “, MK,2002.

20. M. Barni, F. Bartolini , “ Watermarking System Engineering”. Marcel Dekker, 2004.

21.V.Gorodetsky ,V.Samoilov, “Simulation-based Exploration of SVD-Based Technique for Hidden Communication by Image Steganography Channel” , LNCS, 2776, pp.349-359, 2003.

22. V. Korzhik et al. “ A performance evaluation of digital watermark under an additive noise attack “, In: Proc. VII Spanish Meeting on Cryptography and Information Security, pp. 451-470, 2002.

23. H. Malvar, D. Florencia, “ Improved spread spectrum: A new modulation technique for robust watermarking”, IEEET Trans. On Signal Processing, 51, pp. 898-905, 2001.

24. Perez-Freire, F Perez-Gonzalez, “ Spread – spectrum vs. quantization – based data hiding: Misconceptions and implications “, In: Proc. SPIE 17th Symposium, 2005.

25.V. Korzhik et al. “Digital watermarking under filtering and additive noise attack condition “. LNCS, 2776, pp. 371-382,2003.

26. V. Korzhik et al. “ The Performance Evaluation of Estimation Attack on Spread – Spectrum Based Watermarking of Images” (submitted).

27. C, Wu et al, “ Robust Audio Watermarking for Copyright Protection”. SPIE’s, 44th Meeting, pp. 18-23, 1999.

28. K.J. P. Liu et al, “ Multimedia Fingerprinting Forensics for Traitor Tracing”, Hindawi PC, 2005.

29. A, J. Menezes et al, “ Handbook of Applied Cryptography”, CRC, 1996.

30. M.H. Lee, V. Korzhik et al, “Image authentication based on modular embedding”, IEICE Trans. Inf. N1, E 89-D (4), pp. 1498-1506, 2006.

31. M. Goljan et al, “ Distortion – free data embedding for images”. Proc. IHW’ 2001, pp. 31-45, LNCS.

32. V. Korzhik et al ,“ Methods of exact image authentication based on watermarking technique “, Problems of information security, N1, 2008, ( in Russian).

33 J. Fridrich et al, “ Invertible Authentication Watermark for JPEG Images”.

34. N. Zhicheng et al, “ Reversible Data Hiding “, IEEE Trans. on Circuits and Systems for video Technology, vol.16 , N3, pp. 354-362, 2006.

35. J. Proakis, “ Digital Communications “, McGraw – Hill, 1995.

36. Gui – jun Nian et al, “ Research of improved echo data hiding : audio watermarking based on reverberation “, In : Proc. ICASSP ‘ 2007 pp. 177-180.

37. N. Cvejic, “ Algorithms for audio watermarking and steganography”, http : // herkules.oulu.fi / isbn9514273842/.

38.A.Davydkin, “Investigation of model-based steganography”, master thesis, SPbGUT, 2009.

39.B.Ryabko, D.Ryabko, “Asymptotically optimal perfect steganographic systems”, Problems of Information Transmission, 2009, vol.45,N 2,pp.184-190.

40.K.Szczypiorski , “Steganography in TCP/IP Networks. State of the art and Proposals of a New System, MICCUPS,

41.J.Fridrich et al., “Efficient Wet Paper Codes”,

42.V.Korzhik et al, “Stegosystem Based on Noisy channels”, IJCSA (submitted).

43.F.MacWilliams,N.Sloane,”The Theory of Error Correcting Codes”,NHPC,1977

44.J.Fridrich, “Steganography in Digital Media.”,Cambridge Press,2010.

The subject matter of the course

No / Title of the theme / Totally on
the theme / Lectures / Exercises / Term
project / Personal
work
1 / Introduction / 2h / 1h / - / - / 1h
2 / SG systems with
Embedding in LSB / 4h / 1h / 1h / 2h
3 / Real SG and SS-based SG / 4h / 2h / 1h / - / 1h
4 / SG for other cover objects / 5h / 2h / 2h / - / 1h
5 / Perfectly secure and almost perfectly secure SG / 6h / 2h / - / - / 4h
6 / Steganography based on noisy channels / 8h / 2h / - / 4h / 2h
7 / Blind steganalysis / 3h / 2h / - / - / 1h
8 / The main applications of digital watermarking / 2h / 1h / 1h
9
10.
11 / Technique of watermarking
Informed encoder
Sophisticated attacks on WM / 10h
2h
4h / 2h
1h
2h / 2h / 4h / 2h
1h
2h
12 / Breaking of WM synchronization.
Collusion attacks. / 6h / 2h / 2h / 2h
13.
14 / Content authentication based on WM systems
Watermarking of audio cover objects / 4h
2h / 2h
1h / - / -
- / 2h
1h
15. / Capacity of SG and WM systems / 2h / 1h / 1h
Totally on the course / 64h / 24h / 8h / 8h / 24h

Remark .Students are free to select any one of two term projects on the theme 6 or 9 presented in Table above.

Distribition of student’s final grade depending on particular grades :

1.Presence at the lectures -10%

2.Exercises-30%.

3.Term project-20%.

4.Current control of the knowledges-10%

5.Final exam-30%.