top | item 19445323

(no title)

imraj96 | 7 years ago

92.7m+ tags!! How did they annotate the images? Was it done manually by humans or a neural network?

discuss

order

gwern|7 years ago

All humans, believe it or not. Never underestimate the power of procrastinating grad students or anime weebs.

At present, the most advanced tagger for Danbooru, DeepDanbooru https://www.reddit.com/r/MachineLearning/comments/akbc11/p_t... , still isn't good enough to do annotation by itself but I think that's mostly because no one has really tried.

bspammer|7 years ago

From the website

> The image boorus are longstanding web databases which host large numbers of images which can be ‘tagged’ or labeled with an arbitrary number of textual descriptions; they were developed for and are most popular among fans of anime, who provide detailed annotations.

i.e. it's pulling from a public website which crowdsources the tagging