7

Visual content ID system with a public API

Project Type
Open source idea
Summary

Visual content ID system with a public API

Description
It's a system where content creators can upload the images they create and claim its copyright. It's also a system capable of checking if an image is copyrighted or not. The images uploaded to the system will not be visible to the public, of course. It's similar to Youtube's Content ID system although the said system is open for all and is only for image contents. It is designed to be integrated into the image upload process on existing social media platforms such as Facebook, Twitter, Tumblr, etc., probably even here on devRant! Most of the time, people just download and then post (not share!) images on their own page or account to gain likes from other people without crediting the owner of the image. Some even monetize their posts!! A lot of content creators face this problem very often. Check out @forexposure_txt on Twitter and you'll probably get the idea of what I am saying. (It's funny though how some people fight over the ownership of memes. 😂) The main goal of this kind of project is to help content creators ensure that... 1) they can completely engage with his/her target audience. 2) they know all the places where his/her contents are posted assuming that he/she allowed other people to use it under a strict condition, e.g. posted or shared with a mention of his/her name and contact details. 3) no one else may use their content without credit to gain popularity (likes, favorites, shares, retweets, reblogs and other forms of social media engagements) and/or to create profit from it. Flows of the major ystem processes: A. Claim copyright for an image 0. User performs the appropriate authentication process. 1. User submits an image to the system. 2. The system checks if the image to be uploaded is not yet owned by another user. a. If true... i. The system saves the image and associates it to the user's account. ii. The system notifies the user that the upload was successful and that the user is now the holder of the uploaded image's copyright. b. If false... i. The system discards the image. ii. The system notifies the user that the upload failed and the image the user is trying to upload is an IP of another user. B. Check if no one is holding an image's copyright Note: The word 'user' in this context may pertain to a person using the web portal or to some other web service like Twitter or Facebook. 0. User performs the appropriate authentication process. 1. User submits an image /using the web portal or using the API. 2. The system checks if the image to be uploaded is not yet an IP of another user. 3. The system sends a response, which indicates if the image is copyrighted or not, back to the user. Flaws regarding the system: If there are flaws on the system that you can think of, please don't hesitate on putting it on the comment section so that everyone interested could discuss it. * I am just posting this here to give y'all some idea. I have no plans to immediately start this kind of project or immediately participate in one. Something like this should not be rushed and should be discussed thoroughly by lots of people. This is complex, big, and very costly. Probably even need some investors. Just thinking of how to quickly and efficiently match similarities between an image and a pool of other images tickles my brain. Eurgh. * I just would like to contribute something that will protect the artists and their works. They're important to our lives you know! * Please forgive me if I misused the words 'copyright' and 'IP' on this post. I am no legal expert and definitely not really a visual content creator. I know a little about software licensing though.
Tech Stack
TBD
Current Team Size
TBD
URL
Comments
  • 4
    I like the project the url :)

    So it would be a like a proof-of-creation thing, interesting!
  • 2
    would be nice to also be used so the creator of memes could get the credit they deserve
  • 0
    I hoped "flows" was a list of "flaws", because there is too many to even address.
  • 0
    @JoshBent If you have the time, will you please tell us why you think those 'flows' are 'flaws'? I do not have a clear view of the whole problem so please share your perspective.
  • 1
    @monsieurloin the flows are not flaws, I instead of a list for "flows" wished for a list of "flaws" as theres so many in that idea
  • 1
    @JoshBent Ohhh I'm sorry. I totally misunderstood what you said.

    Is it okay for you to give us an example? To be honest, I can't think of one.
  • 0
    @monsieurloin

    - People would have to actually use the system either directly or indirectly, that leads to following issues:

    -- If your API would be turned into some sort of website, it wouldn't be used, since people first of all wouldnt even know it exists, nor actively use it, because of multiple factors.
    -- All major companies would have to use your system licensed in their website, which most likely won't happen, because:
    --- You named for example facebook - you are talking about the same facebook here, that is one of the main runners in AI and all sorts of image detection algorithms - you not only want to beat that huge developer base they have dedicated to AI, but also imply that they would acquire it, instead of just recreating the same system - better - in 30 minutes of their lunchbreak
  • 0
    @monsieurloin
    - There will be a point where actual copyright holders won't be able to claim copyright, because somebody else uploaded their work before them, leading to problems on who actually made the image and who has the rights of it

    - Your image detection would have to be incredibly advanced, to a point where a small developer team just can't keep up anymore, for example:
    -- how do you handle images that just got slightly (not just a pixel, but smudged parts for example) or destructively modified? - you would need a comparison detection more advanced than what OpenCV can do (not just hash comparison), so you would have to roll your own, which can handle huge gaps between modifications
  • 0
    @monsieurloin
    - The system deployed on youtube by google (which would just monetize their system to others if that area gets remotely popular, with an already developed algorithm) - works, because they have total monopoly, nobody cares some vimeo, youtube is the only platform where people upload videos, both copyright holders and the copycats, leading to an easy game for google to filter infringments
  • 0
    @monsieurloin one could go on and on, thats just some things coming to mind immediately
Add Comment