We integrated Skale ML to validate the content (video or image) participants submitted to AnonHero. The idea behind this was using the ML system to receive a proof of content, using a model we trained to validate whether an image had a building or not.
If we didn't have the time constraints, we could have trained a model and used data that validated the content further, tying it to a specific location and further validating if the content was real.
Before a solution like this goes to production I would image the biggest challenge will be getting processing time down.
Log in or sign up for Devpost to join the conversation.