Guide 29 - Tutorials | Articles

Google is very serious about respecting the rights of video content owners, especially when it comes to copyright infringement. But if we consider that every minute is up 20 hours of videos uploaded, how does YouTube know and remove videos with copyright infringement content.


In the following video, Margaret Stewart, head of user experience team from Youtube, comments on how the copyright proof system work against plagiarism:

For example, as soon if you upload a music video, the identification system will automatically detect if that content is copyright or not, advising and giving options to the record company what to do with the video, delete it or not. But how does this system work in the background ?

1. First, the owner of a video or the content copyright introduces the characteristics of the video in the YouTube database, along with user policies that say what to do with the content if it is copied.
2.  While uploading a video, youtube software automatically compares every parts of the video with their database. The heatmap of the photograph shows the comparative reference file with the file uploaded by the user.
   
3. The system checks a video every time to see if they match. It is capable of detecting even slowdown or distort sounding even clips.
4. When a copy is detected, it automatically applies the policies set by the owner.
   

And all of this is done with more than 100 years of video being uploaded each day and their system compares them with millions of reference file in their database. I think the interesting thing is not the subject of copyright, but the amazing ability of their systems and algorithms that do this almost instantly.

BTW, the wedding song you saw in the video can be found here. (too good) What do you think ?