Open robertleeplummerjr opened 7 years ago
Fun, but very hard! I think the way to go on this is to split the big image into smaller chunks (say 100x100 each) and then run a convolutional net over each one of those chunks. Let it classify if waldo is present. A generator is a whole other story though...
Agreed! Progress for the underlying library requirements are here: https://github.com/harthur-org/brain.js/issues/39
@robertleeplummerjr were you able to achieve this?
The entire of brain.js v2, which we hope to have a release candidate out by the end of the month, will be the catalyst for this, and I have been working on it for essentially a couple years. I wanted to not just do it, but make it so easy a child could do it.
I was told:
Fun, but very hard!
I want to tell my future self, future @wagenaartje, and future @samparsky: "ha, easy"
We will see, but I know the api, is working, and will give us the wings to achieve this.
I was wondering if the where's waldo generator was completed?
That would be very cool
This came to me talking about the neural net when explaining it to a friend, might be fun.