Abstract
Forensic watermarking is often used to enable the tracing of digital pirates that leak copyright-protected videos. However, existing watermarking methods have a limited robustness and may be vulnerable to targeted attacks. Our previous work proposed a fallback detection method that uses secondary watermarks rather than the primary watermarks embedded by existing methods. However, the previously proposed fallback method is slow and requires access to all watermarked videos. This paper proposes to make the fallback watermark detection method faster using perceptual hashes instead of uncompressed secondary watermark signals. These perceptual hashes can be calculated prior to detection, such that the actual detection process is sped up with a factor of approximately 26,000 to 92,000. In this way, the proposed method tackles the main criticism about practical usability of the slow fallback method. The fast detection comes at the cost of a modest decrease in robustness, although the fast fallback detection method can still outperform the existing primary watermark method. In conclusion, the proposed method enables fast and more robust detection of watermarks that were embedded by existing watermarking methods.
Highlights
Forensic watermarking is used to track down digital pirates after they leak video content, i.e., to perform traitor tracing [1,2,3], i.e., each user receives a uniquely watermarked version of the video
The secondary watermark is exploited in the fallback detection method to improve the robustness of existing watermarking techniques, i.e., the secondary watermark of a leaked video is compared to the secondary watermarks of all watermarked versions of the video
This section experimentally evaluates the proposed fallback method based on perceptual hashes of secondary watermarks
Summary
Forensic watermarking is used to track down digital pirates after they leak video content, i.e., to perform traitor tracing [1,2,3], i.e., each user receives a uniquely watermarked version of the video. The previously proposed fallback watermarking method effectively improves the robustness of existing techniques using secondary watermarks, the detection process is non-blind, i.e., a leaked video (segment) is compared to all distributed watermarked videos. This means that access to all watermarked videos is required for fallback detection. To alternatively use uncompressed-domain watermarking methods in a scalable way, they can be applied on each individual client’s device, after decompressing the received video [15,16] Such techniques decrease the overall security of the system, since the unwatermarked video will be temporarily stored on the client’s device, and can potentially be accessed by malicious users. Targeted attacks may be invented that delete a watermark with less impact on video quality
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
