Why we invested in Vestigit
The Covid pandemic has significantly increased the demand for content streamed on the Internet. This should be a good sign for video content and live stream providers. In fact, this situation does not translate into their sales results as much as it might seem. The reason for this is widespread piracy, that is, illegal sharing and viewing of content from which its creators and distributors do not derive any benefit. Therefore, authors and distributors are looking for solutions that will enable them to fight illegal distribution of their works. One of such solutions is provided by Vestigit, in which we, RKKVC, have just invested.
Background
The pandemic has resulted in people who have previously been away from home either at work or elsewhere, being locked at home. Many ways of entertainment and social engagement have been lost. This made them more willing than before to reach for movies and series as well as other content available on the Internet. Only in the USA, in March 2020, i.e. at the beginning of the pandemic, the amount of content streamed increased by 73% compared to the previous month [1]. Today, despite the fact that we are dealing with the pandemic better and better, and further entertainment “outside the home” is released from the restrictions, we are not resigning from the services of Internet video content providers. Video streaming has entered our lives and has become something regular for us.
The problem is that, to a large extent, some of the beneficiaries of this boom in video streaming are not content creators, but illegal distributors. While RKKVC is bold in computer vision and Image Recognition, we received a lot of feedback that piracy and theft on streaming market is growing rapidly. It is estimated that up to 24% of global Internet bandwidth is used for illegal video streaming. According to some estimates, the TV industry could lose as much as $97 billion a year.[2]
Therefore, it should come as no surprise that content providers are looking for every way to counter piracy. One of such solutions is offered by the Polish company Vestigit, in which RKKVC invested at the beginning of July as a co-lead with Redge Technologies — the CEE industry OTT provider[3] and with the support of Custodia Capital.
Vestigit is a Polish startup dealing with the protection of video content provided under, inter alia, VOD platforms by identifying illegally retransmitting recipients. The problem of the lost income of video creators and distributors was noticed in 2018 by an experienced team of scientists working at the Wrocław University of Technology, dealing with cryptography of neural networks and traders with over 15 years of experience in OTT and IPTV transmission. It is agreed that this is one of the best Uni in CEE that develops advanced algorithms in that sector.
Vestigit provides a cybersecurity platform that protects video content by identifying illegal sharing through automated implementation of end-user specific watermarks in real time. This platform is used to mark video content using the watermarking technique (adding an invisible for a human eye watermark) in an individualized way to the recipient of the signal, as well as to effectively detect illegally distributed video materials and inform content provider whis user is violating the terms in no time. Optimized algorithms enable the analysis of the tested material in close to real time and a quick search of the internal database of historical content. One of the distinctive solutions used by Vestigit is an advanced web-crawler, which monitors places on the web where videos infringing copyright may be shared. The product is dedicated to all visual materials available on websites.
Do you see the difference?
As RKKVC, we see great potential for using Vestigit technology in the video streaming industry. The company’s development so far, including cooperation with leading suppliers of video content in Poland at the Proof of Concept level, convinced us that the company is able to respond to the growing demand of TV suppliers for services that will protect their content against piracy.
WHY: Technology
Vestigit uses advanced techniques that combine image processing algorithms, cryptographic techniques and steganographic methods. Some modules were built on the basis of the latest artificial intelligence techniques, in particular the so-called adversarial neural networks (GANs) modified for the purpose of the project.
Vestigit technology uses an algorithm for embedding and extracting a watermark based on a series of fast, non-linear arithmetic operations. The platform uses AI / ML solutions, which enables a real and scalable increase in resistance to pirate attacks with the possibility of extending the monitored attack classes.
The main characteristics of the solution include:
· Personalizing the video stream and associating it with a specific end user;
· Requiring only a short excerpt of video material to identify a specific user;
· Ability to identify illegal streaming of live broadcasts;
· High resistance to actions aimed at preventing identification of material distributed in violation of the law;
· Interoperability — embedding a watermark is based on the so-called A / B coding;
· Support for the full range of profiles used, including 4K;
· Possibility of securing content for various scenarios — both for on-line transmission and materials available for a long time in the broadcaster’s offer.
Let’s dig deeper
The product consists of two systems. The first one is used to apply and detect unique characters for each transmitted video material. This is meant to enable the identification of the user for whom the copy of the film was prepared. The application works in complex way:
• it is difficult to remove the entered characters without reducing the quality of the film to such a level that it becomes worthless to the potential viewer (immunity);
• the embedded watermark does not significantly affect the reception of the original video content (transparency). What is more, it is almost impossible to detect the mark with an human eye at any image resolutions.
Vestigit functionality is implemented using the mechanism based on the fact that on OTT websites movies are typically sent as a sequence of short fragments called “chunks”. In the proposed solution, the streaming server prepares two versions of the film — A and B, and as a result, each of the chunks comes in two versions. Then, for each HTTP session, a unique sequence of chunks A or B will be generated, based on which it will be possible to identify the end-user. The construction of the sequence, i.e. the mixing of the A and B chunks is carried out implicitly and the OTT manifesto downloaded by the player and, what is important stays constant at all time. The identification consists of recognizing whether subsequent fragments are of type A or B. On this basis, the coded sequence is played, which is the unique identifier of the session and, as a consequence, clearly indicates the user. This functionality is implemented by a mixing module that creates the appropriate manifests that encode the key for a specific user.
The core company’s product called Vestigia is a software for marking video content with watermarking technique (adding an individualized watermark) for the signal recipient as well as for effective detection of illegally distributed video materials, even in the real time. The product uses advanced techniques combining image processing algorithms, cryptographic techniques, and steganographic methods. Some modules were built based on the latest techniques of artificial intelligence, in particular, the so-called adversarial neural networks (GAN) modified for the project’s purposes. To the knowledge of the contractors, these solutions are new on a global scale. Advanced classic algorithms (e.g. frequency domain labeling) have also been used after adapting them to the project’s technological requirements. The methods used are designed to ensure the irremovability of the information added, the possibility of its extraction, and, as a result, the identification of the infringing user, both during and after the transmission.
Therefore, it can be said that that Vestigia core has two main pillars. The first is used to apply and detect the watermark, while the second is to index historical content and indicate whether the video material comes from the indexed collection, which is supposed to indicate copyright infringement. Optimized algorithms enable the analysis of the tested material in near real-time and a quick search of the historical content database. The product is dedicated to all visual materials transmitted by cable television or made available on OTT platforms.
The product also includes a detection system that uses extensive and effective mechanisms for preliminary indexation, thanks to which it is possible to very quickly analyze materials in terms of similarity to the collected historical data with a low probability of incorrect classification.
The modular architecture of the system allows it to be quickly adapted to various requirements, in particular regarding the level of security and the specificity of video materials to be protected. Applied, possible effective algorithms provide high scalability and support for a large number of users at the same time, as well as protection of a large volume of historical video content.
It should be noted that the coding described above (referred to as A / B coding) was used in systems performing similar tasks. Modules that embed watermarks to distinguish between versions A and B of the film are a unique implementation. On the other hand, versions A and B are (almost) indistinguishable from the original film. To this end, a specially prepared product module currently allows the embedding of two types of watermarks — one based on transformations in the frequency domain the other was created based on GAN adversarial neural networks. Other modules operating within this system are used to detect the coded sequence based on a fragment of video material.
The second system is used to identify the film based on its fragment. The system works as follows: to quickly compare films, an advanced, extensive pre-indexing mechanism is used. The films collected in the database are classified employing very fast, light “functions” classifiers. To determine whether a given video material is part of a movie from the historical content database, first, fast preliminary classifiers are used. In the second stage, the tested material is compared with a very limited number of films contained in the historical database with appropriate values for the group of classifiers used.
WHY: Team and Partners
One of the things we always pay attention to in RKKVC is the founders team. The team experience is undoubtedly a strong point at Vestigit. Both Rafał Radawiec, CEO, and Krzysztof Kołodziejczyk, CTO, have a strong academic background in the field of cybersecurity from the years of work at the Wrocław University of Technology. Additionally, Rafał has experience in this area gained while working for IBM. For us, as investors, such a background guarantees that the team is able to bring the project to completion.
An additional factor influencing our decision was the cooperation of Vestigit with Redge Technologies (previously Atende Software), a leading provider of OTT services in CEE. We noticed large synergies that comes from Redge Technologies with Vestigit. Our partner can support Vestigit in variety of ways. From R&D with their technological advancement, hardware and commercially to offering Vestigit products to its clients..
WHY: Market
Video streaming is one of the most popular ways to use the Internet today. It is estimated that by 2022 it may be responsible for up to 82% of global internet traffic. It is worth noting that over 40% of Americans and over 20% of Europeans today use paid video content.
The global OTT streaming market has grown from $ 104.11 billion in 2019 to $ 161.37 billion in 2020 at an astonishing compound annual growth rate (CAGR) of 55.0%. Considering the current trends, we can expect the growth rate to be maintained.
At the same time, the losses of TV content creators due to piracy are growing. It is estimated that piracy and account sharing will cost streaming video providers $ 12.5 billion in 2024, an increase of 38% over 2019’s loss of $ 9.1 billion.
We see this as a great opportunity for Vestigit, whose tool can effectively counter piracy by allowing illegal distributors to be traced, end user identification, and even blocking the transmission of content in the case of live broadcasts. We believe that this solution will revolutionize the so far unequal struggle between video streaming providers and pirates.
There is more!
In addition to all of the above, the company posses another project which is called Adspective. Adspective is a platform that automates the creation of high-quality advertising in multimedia content using AI. It allows ad networks and video distributors to add ads directly to the content of existing videos and photos, thereby increasing the available online advertising space.
Hints for the future founders
When thinking about starting a company, we often consider entering industries that are popular and most often targeted at mass customers. Vestigit is an example of a company that took a different path. They created a product dedicated to a narrow group of B2B customers in the TV industry, which, however, perfectly addresses their current needs and solves the problem that generates large financial losses for these customers. The creators of Vestigit showed that you do not have to follow trends and do what everyone else does, but it is worth focusing on a specific and pressing problem of a specific customer group.
Vestigit is also an example of a company that was able to finance its activities without the participation of investors in the initial stage of its existence, thus not disposing of its shares too early. The company financed itself through bootstrapping, EU grants and loans.
As RKKVC, we appreciate founders who do not limit themselves to one source of financing and, if necessary, have the courage to take loans which always involves some risk. Especially the latter testifies to their belief in what they are doing. In such a situation, we are willing to reward founders by supporting them in settling their loans and, if necessary, in fixing their cap table.
[1] How Covid-19 is changing online streaming behavior globally — NPAW
[2] All about Video Piracy & How it can be stopped | VdoCipher Blog
[3] OTT — An over-the-top (OTT) media service is a media service offered directly to viewers via the Internet. OTT bypasses cable, broadcast, and satellite television platforms, the types of companies which traditionally act as controllers or distributors of such content.
