Haven’t Done It Before? Don’t Claim It Is Easy Then.
A couple of weeks ago during lunch, myself and a group of colleagues began brainstorming ideas for an upcoming hackathon. One of the ideas I proposed was doing a web video editor similar to the one Youtube has.
The discussion about the web video editor stumbled into a side conversation about detecting copyrighted material in the video material being uploaded for the editor.
The discussion about identifying copyrighted material lead to an individual in the group making the claim, ‘It is easy to achieve this.’
There is a lot of back & forth; I try to shed some light on the complexity of the problem because I entirely disagree with the claim that this is a trivial problem to solve.
I notice the side conversation has hijacked our original discussion, and I ditch my attempts to explain the complexities around the problem. The person summarizes their final solution before moving along, ‘Just take small clips throughout the larger clip and just compare the sound with a database of copyrighted material.’
Talk Is Cheap. Doing isn’t.
I haven’t had the interest or opportunity to attempt solving the problems around detecting copyrighted content, and this is the reason at this moment I would never make a claim that it is an easy problem to solve.
Here are just a handful of issues around the proposed solution:
- What if the video clip has no sound?
- What do you do about sound distortion from different compressions by the user?
- Where is this original database?
- What if a user applies a new soundtrack but the video is copyrighted?
Detecting copyrighted content is not a trivial problem, and this is why only a handful of specialized teams have solved this well.
Don’t claim a problem is easy without having solved the problem prior.