An Attempt to Improve Reverse Image Search
An Attempt to Improve Reverse Image Search
Hey guys, This is a demo of a bit of research I did that didn’t really go anywhere but I thought it was interesting so I made a demo out of it. I wanted to improve reverse image search on edited images and photobashed images. I thought it was neat that the algorithm only uses highschool math and the global minimum is found analytically. I did a demo a few years ago on similar work which made it to the hm front page here: https://news.ycombinator.com/item?id=14973741 This demo was hard to write. It was really hard to not to either over or under explain the algorithm. For some this will be a really simple algorithm that could be summed up in a few sentences and for others it’ll require remembering some math they haven’t seen since college or highschool. It gave me an even bigger appreciation for how well YouTubers like 3blue1brown can explain mathematical concepts. Hope you guys enjoy it
Share cardActual performance
Launch Intel predictions
Analyze your own launch →Incorrect prediction on native model
Similar products
HappyCat – A reverse image search for adoptable cats
The first and only reverse image search tool for SEOs
Calcuchord – Reverse search for musicians and songwriters
Reverse Image Search with Facial Recognition
Transformation Invariant Reverse Image Search
PleasantFish (Reverse of Glassdoor)
Dingwings – Reverse Wingdings
Parsing a Reverse Polish Expression
HCaptcha Reverse Engineered
BACK – 2015 in reverse