Explain fundamental watermarking properties that help OTT industry in identifying sources of piracy

Content producers and the over-the-top (OTT) platforms, like Netflix and Amazon Prime, bank on digital rights management (DRM) and forensic-watermarking technologies to control and detect piracy of their premium content. Forensic watermarking has proved to be effective in tracing the source of piracy and producing legal evidence for breaking illegal distribution networks. Below, we discuss fundamental properties that make forensic video watermaks effective and help a seamless and secure viewing experience:

• Perceptual similarity: This feature is also known as imperceptibility and essentially means that a watermarked content file should be sufficiently similar to its un-watermarked version. The distortions caused by watermark embedding and the introduction of perceptible artifacts into the video asset can help attackers in tampering with the watermark information.
• Visibility: Visible watermarking solutions embed a watermark (typically company logo, icon, or courtesy tags) that appears noticeably on DRM protected content. The level of visibility can be controlled depending on the application so as not to hamper the viewing experience.
• Blindness: This feature is usually used to categorize a detection process that is independent of the original information or its derivatives. If the detection process requires only the secret watermarking key, it is known as blind watermarking. In case the detection process requires the original content copy, the watermark, and the secret key, it is called non-blind or informed. Semi-blind approaches require the watermark and secret key.
• Invertibility: Also called reversibility or losslessness, invertibility in a watermarking solution allows the watermarked content to be restored to its original version where there is no embedding distortion.
• Robustness: It can be defined as the ability to detect the watermark and extract the watermarking information after it has been altered. In other words, it is the resistance offered by a video watermarking scheme to modifications in the host signal either due to signal processing techniques or tampering attempts by attackers.
• Embedding capacity: This is the amount of information, usually expressed in the number of bits/sample, that can be stored in the watermark. The capacity or digital payload of the watermark is mutually dependent on other factors, such as robustness and perceptual similarity.
• Error probability: This helps in determining the reliability of a watermarking approach. Some commonly used measures of error probability are bit error rate (BER), false positive rate (FPR), and false negative rate (FNR).
• Security: It can often be confused and overlap with robustness. However, the two features need to be considered separately. While robustness provides resistance against processing techniques and alterations to the watermark, for the overall security of a solution, all possible attacks, such as elimination attack, masking attack, collusion attack, distortion attack, copy attack, etc., need to be analyzed.

Content producing companies need to keep these aspects in mind while choosing the right forensic-watermarking service provider, since their ability to identify the source of leakage and ability to take legal action against it depends on these factors.