Quote from: caroleshaw on November 16, 2017, 03:02:27 PM

Quote from: shogun47 on November 16, 2017, 01:12:35 PM

Quote from: stockblock on November 13, 2017, 03:52:57 PM

Quote from: shogun47 on November 13, 2017, 02:03:45 PM

Quote from: stockblock on November 13, 2017, 04:11:06 AM

Quote from: shogun47 on November 12, 2017, 11:49:41 AM

Quote from: stockblock on November 08, 2017, 11:42:08 AM

Quote from: shogun47 on November 08, 2017, 10:42:17 AM

Hey, OP is Official Post. First page of the thread first post so to say. Your post explains very well where you are going with this project. And your example about the artists makes a lot of sense. How are you going to achieve reliable tracking? What if I take a screenshot of an image and not download the image itself, how are you going to recognize someone abuses the image? Is that possible?

Hi,

The system uses perceptual hashing to identify similar images. Please check our algorithm test page: https://stockblock.io/fingerprint/

The system uses perceptual hash/ Fingerprint to identify an image. The system can identify images even after the image has been seriously altered. The similarity score defines how similar the image is. So similarity score above 90 is the same image.

This fingerprint is used to find duplicates in the platform itself. The same fingerprint is used to find/monitor the images on the internet. The monitoring system will be a background process. The system will crawl the internet and find images and report the same to the photographers monitoring dashboard. The photographer can decide what actions need to take regarding the infringement.

It sounds extremely interesting. Is your algorithm already working and has its success rate been proven? If that thing gets to work properly, you'll provide a real weapon against ownership infringement.

Yes. The algorithm works and detects plagiarism. The same user can upload different versions of the image but it is flagged if some other user tries to upload someone else's image. The system bans the fraudulent user for ever.

This system is a real weapon for photographers. It protects them from copyright infringement.

Ok but banning a fraudulent user is one as effective as the actual barrier to entry the platform. If I can make a million accounts banning is not a real solution to the problem. Does your platform also allow for commerce of copyrights? Like, when I want to use a picture I could send some crypto to the original owner and get a certificate or something like that?

The system automatically checks for plagiarism and bans the user. To keep fraudsters away we plan to keep the system invite only. Unless you have invite, you can not join the network.

The marketplace platform is for the artist. They can sell their artwork with different editions. Single bulk etc. Everything works using ethereum SmartContract. Since the system works on SmartContract, the photographer gets paid instantly. A small percentage fee will be charged for the services and it will be defined in the SmartContract.

The entire ecosystem works on xmc token.

Byers buy artwork using the tokens. Sellers receive tokens.

Smart consideration, I like that with the invite only approach. But it will hinder you a little bit with fast adoption I think. There is a trade-off but it might still be very successful because the quality of your participants should be higher in the end. We'll find out once your platform is operational

I am an active shutterstock member. I have seen fraudsters uploading someone else's images. Shutter stock doesn't verify the authenticity of the uploader. I like stockblocks approach. This project provides more value compared to any stock photography website. None of them provide copyright protection. The just want to take your 70-80% of your earnings and pay pennies. A lot of photographers are fed up.

I hope this project takes off soon.

I am always open for others to convince me and you put forward a very good point. So yes this approach is most likely the right approach to get a high quality network going right from the start.