TU BRAUNSCHWEIG

Seminar Multimedia Security

Overview

Semester: Winter 2016/2017
Course type: Block Seminar
Lecturer: Prof. Dr. Konrad Rieck
Audience: Computer Science Master
Credits: 5 ECTS
Hours: 2
Capacity: max. 8 Students
Room: BRICS 107/108

Schedule

 Date  Step    
 18.10. 15 - 16:30 h  Primer on Multimedia Security, assignment of topics    
 14.11 - 18.11.  Arrange appointment with assistant    
 12.12. 14.12.  Submit paper proposal (Extended to 14.12)
   
 19.12. 20.12.  Submit review of two fellow students (Extended to 20.12)
   
 08.01.  Submit final paper (No Extension)
   
 17.01. 15 - 18:30 h  Presentation with Pizza    

Description

Digital media like images, videos or audio files are an ubiquitous part of our daily life. However, their modification or the creation of copies is nowadays not a big problem. As a result, how sure can you be that the video in the TV news is not manipulated? Has someone added or modified content? There is clearly 'more to a picture than meets the eye'. This seminar deals with the security of digital media. We focus on Digital Image Forensics and Digital Rights Management. You learn how to detect image modifications or to link media to the device that created it. You also learn how to protect or attack the marking of media that protects the authorship or proves the authenticity.

Requirements

The seminar is organized like a real academic conference. You need to prepare a written paper (german or english) about the selected topic in IEEE double-column style. After submitting your paper at our conference system, you write two short reviews about your fellow students' submitted paper to give them feedback how to improve their paper. Then, you will have time to improve your own final paper. Finally, you need to give a talk of 20-25 minutes on the topic. We will also provide drinks and pizza to enjoy the talks at our small MMSec conference.

Contact

The seminar is organized by the Institute of System Security. For questions and further details, please contact

Seminar Topics

Digital Rights Management

Media services like Netflix or Spotify protect their digital media with DRM techniques to prevent pirate copies. This paper gives an overview what DRM techniques exist and which approaches exist to break them.

Digital watermarks

Digital watermarks are a specific DRM technique that allow the invisible marking of media. This enables e.g. the identification of the the cinema where an illegal copy has been filmed. If Eve has access to a watermark detector, she can iteratively change the media and remove the watermark gradually. This paper explores this so-called sensitivity attack and reviews possible countermeasures.

Audio fingerprinting

Audio recognition services like Shazam allow us to identify a music song from a radio for example. For doing so, each music song is fingerprinted. This paper compares the different approaches and implements one of them.

JPEG compression

JPEG is one of the most popular image compression formats. Revealing the compression history of an image might give clues about its authenticity or originality. This paper analyzes the different ways to find out the compression history, even for high-quality JPEG images.

Sensor fingerprints

Our mobile devices are full of hardware sensors which leave an unique fingerprint due to manufacturing imperfections. This enables an attacker to track a device even after a factory reset. This work compares the different sensor fingerprints of mobile devices.

Camera fingerprints

An image can be linked to a specific camera due to the camera sensor fingerprint. While this forensic technique may help in identifying fraud, a reporter however needs a way to publish sensitive images without tracing these images back to her/him. This paper examines the different image anonymization methods against camera fingerprints.

Fingerprint-copy attack

Camera fingerprints can be estimated from a number of images taken with the same camera. As a result, Eve can also pretend that an arbitrary image was taken by Alice's camera by estimating Alice's fingerprint and planting it on her own spurious image. This paper compares different methods to detect such a fingerprint-copy attack.

Denoising algorithms

The heart of many multimedia security approaches are efficient denoising filters that remove noise from images for example. This paper presents the current state of the art of these denoising filters.

References

  • Cox et al. Digital Watermarking and Steganography. Morgan Kaufmann, 2007.
  • Sencar and Memon, eds. Digital Image Forensics. There is more to a picture than meets the eye. Springer Science & Business Media, 2012.

  aktualisiert am 30.05.2017
TU_Icon_E_Mail_1_17x17_RGB Zum Seitenanfang