Dmitriy Yeremeyev , a co-founder of the InvariVision project, spoke about the new frame-by-frame video recognition technology have been developed by the company. What is its trait and why is it better than the Fingerprint technology that lies in a foundation of YouTube’s Content ID?
One of the most controversial copyright law will take effect in 2019. Article 13 of the new EU Copyright Directive, first published in June 2018, made people panic about the future death of YouTube and the ban of memes. What will change in the copyright protection on the Internet, and what do major players think about this?
Almost from the very beginning, YouTube was one of the most popular platforms for pirated video content. When YouTube creators gave the opportunity to bloggers to get paid for displaying ads, the issue of copyright became even more acute. That is why the Content ID algorithm was created.
There are no people in the world who has never used pirated content – be it an unlicensed version of Photoshop, a movie uploaded on torrents or Vkontakte music. For those authors, for whom their creativity is primarily a business, piracy is real harm, because of which they lost half of their profits every year.
We have added 3 days archive for each TV channel. It’s much easier to check how recognition works and what was shown on a channel before and after recognized films. The archive can be accessed for each channel from the list of channels.
We have added to our site new visualization mode – timeline.
Select it in “Show results as” on Results page.
We have improved video recognition quality. Performance is x5 faster and accuracy of recognition is 19% better. This became possible due to specialization for TV sources of our generic all-purpose image recognition algorithm.
Each TV channel have several backup sources now. System will automatically switch between sources in case of translation problems.
We’ve added ability to export results of film recognition to CSV document that could be opened later in Excel.