AI-powered
podcast player
Listen to all your favourite podcasts with AI-powered features
How to Build an Image Recognition Data Set?
Fatht, and i'm curious, you want to build a data set, an image recognition data set, with 80 thousand possible labels. And compare that to existing data sets. I mean, celtic one, o one had a hundred and one classifacation labels. That's very different from 80 thousand. So clearly it's going to be a different kind of scale effort to pull this off. How do you do it? Sometimes i look back and i think, if i knew all the problem we would roll into, i don't know if we would have hhertetrinawav of naivety that drove us into it.