While making a video sharing website (like Youtube), how can we ensure that no one post an obscene video as well as audio content. 2 of the methods could be
1. Using a software to mask skin show. But this wont be able to monitor unwanted audio content.
2. Team of people reviewing each and every video posted. This will be costly and as the hits grow, the feasibility will be less.
Just looking for some ideas on this front.
Nishant Agrawal
Give the users incentives to report content, like community standing. Be responsive in dealing with such reports.
Puneet Sharma
I agree. There is no perfect way for this.
Obscenity has no S.I. unit. What may look good to me may be Obscene for some.
Example – My Dad thinks these DJ Doll type remix videos must be banned, but not me.
Kaushal Bhavsar
There’s a way to do it by a program. I had made a prototype called “safe eyes” which identified skin show but I dumped the project because I thought no one would want it 😛 The best way to start is to write a routine that will identify the percentage of skin visible on screen.
Ankur
Kaushal, there are skin software available but my issue is also related to obscene voice content (which might be there without skin show).
Kaushal Bhavsar
obscene voice content can also be filtered but there are chances of false positives, probably that will affect quality of content. It is tough to identify if a moan is obscene or a signature of pain.
Ankur
And what about words which are clear like abusive words.
Kaushal Bhavsar
That can be possible, and this article is a good starting point.
Ankur
Kaushal, is this actually possible through use of technology?
Kaushal Bhavsar
Ofcourse, it is possible, otherwise I wouldn’t have mentioned it. There might be some shady bearded guy writing this code in his basement by the time we are discussing this.