This year, the ACM MM&SEC Workshop committee is proud to be hosting a series of invited talks on Bridging Research and Reality. These talks will come from relevant experts in the fields of Steganography/Steganalysis, Watermarking, and Multimedia Forensics and will serve to frame the discourse for related sessions within the workshop. Presentation topics will be focused on real-world lessons learned, the challenges of bringing technology to fruition, as well as applications and experiences with MM&SEC technology applications that range from commercial to law enforcement. It is our hope that these sessions will provoke and inspire attendees.
Steganography/Steganalysis Invited speaker:
Chet Hosmer – Chief Scientist & Sr. Vice President, Allen Corporation
Biography: Chet Hosmer is the Chief Scientist / Sr Vice President at Allen Corporation of America and was the Founder of WetStone Technologies, Inc. Chet has been researching and developing technologies and training surrounding steganography and watermarking for over a decade, acting as principle investigator on multiple data hiding research programs. He has made numerous appearances to discuss the threat steganography poses, including National Public Radio’s Kojo Nnamdi show, ABC’s Primetime Thursday, NHK Japan, CrimeCrime TechTV and ABC News Australia. He has also been a frequent contributor to technical and news stories relating to steganography and has been interviewed and quoted by IEEE, The New York Times, The Washington Post, Government Computer News, Salon.com and Wired Magazine.
Hosmer holds a B.S. in Computer Science from the Syracuse University at Utica College. He has 12 journal publications, dozens of conference publications, and has delivered keynote and plenary talks on various cyber security related topics around the world every year.
Topic: Chet Hosmer will be presenting a talk titled “Erasing the Myth – A Case Study of Actual Steganography and Data Hiding Incidents.”
Watermarking Invited Speaker:
Tony Rodriguez – Chief Technology Officer, Digimarc
Biography: Tony Rodriguez joined Digimarc in 1996 and currently serves as Chief Technology Officer. Rodriguez has 25years experience in computer science and image processing research and development. Over his fifteen years with Digimarc, Rodriguez has held senior software engineering and research positions, focused on the development and application of digital watermarking and other content identification technologies. Rodriguez played an integral role in the early development of solutions for identifying and tracking digital images and later in the company’s work with an international consortium of central banks to develop a system to deter PC-based counterfeiting of banknotes. He is currently focused on the integration of content recognition technologies with machine learning strategies to create and manage context on the mobile platform in support of applications from Digimarc and it’s partners that enable users to identify and interact with television, audio and printed content.
Rodriguez is the inventor on 41 issued U.S. patents and 97 pending patents. He is also the author of several published papers on the topic of Digital Watermarking and a chapter in the book Multimedia Security Handbook, published in 2005. Before joining Digimarc, Rodriguez worked at the Intel Architecture Lab as a senior software engineer, focused on video segmentation and streaming technologies. Prior to that, Rodriguez held a variety of software development and engineering positions at Raytheon, Jet Propulsion Laboratory (JPL) and IBM. Rodriguez has a bachelor degree in electrical engineering from the University of Washington and an Executive MBA from Stanford.
Topic: Digital Watermarking exists in the space between human and machine perception and its efficacy is typically measured using application specific economic models. Commercial adoption is predicated on successfully navigating this space and maintaining a clear focus on what the customer’s definition of success is. To do so requires the integration of technologies and models from disparate fields and an acute understanding of workflow. Across a broad cross section of deployments and customers a common set of lessons have been learned and technical challenges identified. Many surround the need to accurately model and document the operational envelope of proposed algorithms and implementations. Doing so, within the economic and workflow constraints of the intended application allows early adopters to advocate for and ultimately facilitate the migration of watermarking research from labs into production. For those technologies that have successfully made the jump into the commercial sphere, there is demand to improve the algorithms and implementations, largely driven by a desire to increase the breadth of applications enabled and to support new platforms such as mobile devices. These improvements typically take the form of growing the constraint triangle (carrying capacity, robustness, imperceptibility) or are related to the underlying implementation, such as computational efficiency/gate reduction, or leveraging platform specific capabilities (GPU, etc.)
The first wave of Digital Watermarking technologies has been successfully commercialized and is now widely deployed in support of industry verticals. The next wave of applications awaiting the research community however is larger and more profound as watermarking enters into the public consciousness as an enabling technology that facilitates new user experiences with media of all types. By sharing the lessons learned and commercial challenges, as we perceive them, our hope is that we can facilitate discussion and opportunities for collaboration.
Digital Forensics Invited Speaker:
Walter Bruehs, FBI Forensic Audio, Video, and Imagery Analysis Unit (FBI/FAVIAU)
Biography: Walter E. Bruehs is employed by the Federal Bureau of Investigation as an Examiner of Questioned Photographic Evidence in the Forensic Audio, Video, and Image Analysis Unit. Walter has testified in both Federal and State courts. In addition to case work, Walter’s responsibilities include actively seeking out and researching emerging digital imaging technologies as they apply to the Forensic arena. He heads a program designed to identify digital cameras, based on sensor noise. He has a Master of Science degree in Electrical Engineering from the University of Maine at Orono, as well as a Bachelor of Science in Electrical Engineering from Clarkson University. Prior to working at the FBI, Walter worked as an Imaging Scientist in the research and development laboratories of the Eastman Kodak Company, where he co-authored a patent, “Method and System for Improving an Image Characteristic Based on Image Content.” In addition to his responsibilities with the FBI he has taught at forensic conferences both domestically as well as internationally and has served as an adjunct professor at Polytechnic University of NYU.
Topic: Bringing new technology into Federal Court. As consumer technology continues to get more sophisticated and digital in nature so then also does the need for the forensic community to develop tools to collect data from these devices. The forensic community is unique in the criteria that it must adhere to in State and Federal courts.
The characterization of a camera and its images using Photo Response Non-Uniformity was recently utilized, at the subsequent trial this analysis passed a Daubert hearing in Federal court. The Daubert criteria define the standard by which new scientific evidence can be admitted into Federal court. This marks a 6 year process to research, develop and ultimately bring new technology into a Federal court proceeding. The project is still progressing and involved collaboration between the Federal government, academia and private industry.