Chengdu, China – October 18th, 2016

Privacy and security support for image data is becoming steadily more important seen the fact that image collections are increasingly more stored in distributed and cloud repositories rather than in private repositories. Moreover, social media and online photo repositories, for example, are currently offering insufficient means to secure privacy-sensitive information carried by the picture or to signal associated IPR metadata. Observing that on a daily basis billions of pictures are shared in JPEG legacy formats on these media, it is evident that embedding additional functionality that would safeguard this type of information and functionality would benefit a significant user base.

Hence, the JPEG Committee has launched a new activity called JPEG Privacy & Security. This activity aims at developing a standard for realizing secure image information sharing which is capable of ensuring privacy, maintaining data integrity, and protecting intellectual property rights. This activity is not only intended to protect private information carried by images - in the image itself or the associated metadata - but also to provide degrees of trust while sharing image content and metadata based on individual preferences. It is necessary to extend the existing coding standards by adding such preferences. JPEG Privacy & Security will explore ways on how to design and implement the necessary functionality without significantly impacting on coding performance while ensuring scalability, interoperability, and forward and backward compatibility with current JPEG standard frameworks.

Proceedings (presentations)

All presentation can be downloaded here.

Program

14h00 - Touradj Ebrahimi (JPEG Convenor, EPFL), “JPEG Privacy and Security - General introduction”

Prof. Touradj Ebrahimi received his M.Sc. and Ph.D., both in Electrical Engineering, from the Swiss Federal Institute of Technology (EPFL), Lausanne, Switzerland, in 1989 and 1992 respectively. In 1993, he was a research engineer at the Corporate Research Laboratories of Sony Corporation in Tokyo, where he conducted research on advanced video compression techniques for storage applications. In 1994, he served as a research consultant at AT&T Bell Laboratories working on very low bitrate video coding. He is currently Professor at EPFL heading its Multimedia Signal Processing Group. He was also adjunct Professor with the Center of Quantifiable Quality of Service at Norwegian University of Science and Technology (NTNU) from 2008 to 2012. Since 2014, he is convener of the JPEG committee.

14h10 - Ambarish Natu, “Summary of Brussels and La Jolla Workshops”

Ambarish Natu currently works as an IT Architect within the Information Architecture Team of the Australian Taxation Office where he is responsible for developing data models for effective compliance activities. Since 2002 Mr. Natu has been involved in the development of several standards under the auspices of the ISO, IEC and ITU-T as Editor/Chair/Co-Chair. Mr. Natu has been an active contributor on Australian Standards Committee IT-029, CT-002 and IT-039 for more than a decade. Mr. Natu is a Fellow of the Institute of Engineers Australia, a Chartered Practising Engineer and a Senior Member of IEEE. He currently serves on the ITEE College Board, The Canberra Division Committee and the ITEE Canberra Division Committee of Engineers Australia. Mr. Natu is also the Vice-Chair and Webmaster for the IEEE ACT section in Australia.

14h20 - Prof. Chengqing Li (Xiangtan Univ.), “When Cryptanalysis Meets Image Processing: the Perspective of a Cryptanalyst”

Security and privacy of image data are serious concern of every person living in the cyber world. To solve the change, a great number of image encryption and protection schemes were proposed in the past two decades. This talk will review the design ideas and security of most typical image encryption schemes from the perspective of cryptanalysis. Representative security problems and the contradictions among the related key factors, such as efficiency, security level, and usability, will be introduced. This talk will also try to clarify the relationship between security and image processing techniques.

Chengqing Li received his M.Sc. degree in Applied Mathematics from Zhejiang University, China in 2005 and his Ph.D. degree in Electronic Engineering from City University of Hong Kong in 2008. Thereafter, he had been working as a Postdoctoral Fellow at The Hong Kong Polytechnic University till September 2010. Then, he joined the College of Information Engineering, Xiangtan University, China as an Associate Professor, where he received his Bachelor degree in Mathematics before. From April 2013 to July 2014, he worked at University of Konstanz, Germany, under the support of the Alexander von Humboldt Foundation. Dr Li focuses on security analysis of image and chaos-based encryption schemes. He has published forty papers on the subject in the past ten years.

14h45 - Dr. Gaëtan Le Guelvouit (b<>com), “Image Encryption During JPEG Compression Process”

While public image platforms ans social networks enable users to share images easily, confidentiality and privacy related to those contents became an important concern. Usually, encryption is the tool of choice for this kind of issue. Nevertheless, classical encryption algorithms cannot be used as this to protect images on sharing services. We propose in this talk an encryption algorithm dedicated to JPEG images, which is designed to be robust to the recompression process applied by most platforms. This first work will be illustrated by the results form a public experiment launched this year.

Gaëtan Le Guelvouit joined the INRIA for a Ph.D. in Computer Science defended in 2003. Then, he worked at Supélec Paris and as a security consultant at Capgemini. Since 2006, he has been at Orange Labs as a research engineer, where he was distinguished as Senior Security Expert. His main research interest is data hiding (watermarking and steganography), i.e. the art of hiding information within digital content. In 2013, he joins b<>com as Manager of the Digital Trust & Identity Lab and works on security issues related to multimedia content. Gaëtan is also one of the co-founder of Lamark, a firm dedicated to copyright protection for photography professionals. 

15h40 - Dr. Ru Zhang (Beijing Univ. of Posts and Telecommunications) “Novel steganography using JPEG-compressed images”

A novel steganography based on joint photographic expert-group (JPEG) is proposed. The proposed method embeds two secret bits into each middle frequency coefficient using the least two-significant bits (2-LSBs). For the rest coefficients, the quantization error, resulting from processing the JPEG-compressed image, is compensated while embedding variable-length secret bits. At last, a JPEG stego-image is generated using the quality factor with the value of 100. Experimental results show that, the proposed steganographic method can provide a high information hiding capacity and keep the visual quality well.

Ru Zhang received Ph.D. from Beijing Institute of Technology, China, in July 2003. In 2003, she joined the School of Computer Science at Beijing University of Posts and Telecommunications, where she is currently an assistant professor. Her interest research fields are information hiding, watermarking,multimedia security. She authors over 60 refereed international journal/conference papers and chapters, and holds 12 patents. She got the National Award of Teaching Achievement, “Teaching standardization and architecture of Information Security Subject ” in 2009.

16h05 - Prof. Rongrong Ni (Beijing Jiaotong Univ.) “Protection and Privacy for JPEG Images”

With the increasingly development of computer science and the Internet, multimedia applications and sharing are becoming attracting. The manner of capturing, storing, processing, utilizing, and sharing the images is changing. People are familiar with searching images by means of the Internet, and downloading useful images for further applications and editing. Meanwhile, digital images are now easily edited or modified by many software(e.g. PhotoShop). This phenomenon may harm the ownership of the digital assets. The intellectual property rights can not be protected appropriately. Privacy related problems are emerging and need to be solved. Although encryption to metadata provides available methods to privacy protection, it still doesn’t pay attentions to the image contents. In this talk, we focus on data hiding and lossless recovery of JPEG codestream. The embedded data can be the corresponding verification information or copyright information. We also introduce the data hiding in the encryption images. At last, we will discuss the privacy protection framework for special objects or some sensitive objects in the images.

Rongrong Ni received the B.S. degree and the Ph.D. degree from Beijing Jiaotong University (BJTU), Beijing, China, in 1998, and 2005, respectively. Since 2005, she has been the faculty of the School of Computer and Information Technology and the Institute of Information Science, BJTU, where she is a Professor since 2013. Her current research interests include image processing, data hiding and digital forensics, pattern recognition, and computer vision. She was selected in the Beijing Science and Technology Stars Projects in 2008, and was awarded the Jeme Tien Yow Special Prize in Science and Technology in 2009. She is the Principal Investigator of four projects funded by the Natural Science Foundation of China. She has participated in the 973 Program, the 863 Program, and international projects. She has published more than 80 papers in academic journals and conferences, and holds six national patents.

16h30 - Dr. Fan Kefeng (China Electronics Standardization Institute) “Advances of data security and privacy standardization”

Fan Kefeng received Ph.D degree in test signal processing from Xidian University, Shaanxi, China, in Dec. 2007. From Feb. 2008 to Nov. 2010, he was doing the postdoctoral in the field of cybersecurity, Beijing University of Posts and Telecommunications, Beijing, China. After Nov. 2010, he worked as a senior engineer in the field of multimedia technology standardization in China Electronics Standardization Institute(CESI), Beijing, China. Since May, 2013, he worked as the deputy director of the Research Center of Cybersecurity of CESI. His current research interests include data security, ICS security, and signal processing. He is member of ISO/IEC JTC1/SC27 and IEC/ACSEC. He was the convener of smart TV SWG of IEC TC100 since 2012. He was an expert of IEC TC100 since 2007. He was appointed as the Young Professional of IEC in 2011.

16h55 - Dr. David McNally (EPFL) “Legacy JPEG compliant transmorphing to preserve privacy in social networks”

Social media networks pose unique privacy challenges. Users who post to such networks effectively relinquish control of the content they choose to share. In the case of images, privacy management poses particular problems: In general, effective privacy protection negatively affects image appearance, restricts spontaneity and is therefore not attractive for most social network users. Substituting privacy sensitive image regions with alternate image content can serve to protect privacy in a visually appealing manner. To realize this, a JPEG compliant technique is presented which allows for the selective transmorphing of image regions. The image content of the transmorphed region is stored within the JPEG file structure in a standards compliant manner. This supports reversal of the transmorphing process and hence removal of the applied privacy protection without access to external data. Protecting the image content of the transmorphed region through cypher-text policy attribute-based encryption allows for the provisioning of both privacy protection and management without the need of a trusted third party. Employing machine learning techniques, privacy protection in images can be automated in a context aware fashion. This has the potential of simplifying the process of protecting privacy and therefore encourage the adoption of such tools by social media users.

David McNally currently works as senior project manager with the Multimedia Signal Processing Laboratory at EPFL (Switzerland). He received his M.Sc. and Ph.D., both in experimental physics, from the Swiss Federal Institute of Technology in Zurich. From 1993 to 1997 he worked at the European Centre for Particle Physics (CERN) developing advanced visualization systems for particle physics analysis. In 1997 he joined EPFL working on computer performance analysis and optimization. In 1999 he co-founded Genimeda SA (Lausanne) developing perceptual quality metrics for mobile radio resource management. In 2004 he founded Photonique SA (Geneva), the first company to sell Silicon Photomultiplier technology globally. In 2010 he joined SensL (Ireland) as Director of marketing and sales and in 2011 he founded CM-Photonics SA, a Sino-Swiss research company developing medical imaging and radiation detector systems.

Proceedings of previous workshops

• 1st workshop, Brussels, Belgium – October 13th, 2015
• 2nd workshop, San Diego (La Jolla), CA, USA – February 23rd, 2016

Organising Committee

• Yao Zhao (China)
• Jinhu Lu (China)
• Ambarish Natu (Australia)
• Takaaki Ishikawa (Japan)
• Peter Schelkens (Belgium)

Referencing

JPEG Privacy and Security 3rd Workshop Proceedings, ISO/IEC JTC1/SC29/WG1, wg1n73012, Chengdu, China, October 18th, 2016.

More information

To stay posted on the action plan for JPEG Privacy & Security, please regularly consult our website at jpeg.org and/or subscribe to our e-mail reflector.