Facebook Rosetta: Machine Learning to Understand Text in Images and Videos

Forum Moderators: not2easy

Message Too Old, No Replies

Facebook Rosetta: Machine Learning to Understand Text in Images and Videos

engine

9:38 am on Sep 12, 2018 (gmt 0)

Facebook is using machine learning to extract text posted in images and videos. The system, it calls Rosetta, extracts texts from billions of images and videos on Facebook and Instagram in real time, and currently on English and Latin character sets, with more languages under development.

It's an interesting article published by Facebook, and it's worth a read, imho.

[code.fb.com...]

keyplyr

10:42 am on Sep 12, 2018 (gmt 0)

Too bad FB isn't using Ex Machina learning. Might be fun to have Alicia Vikander scan my memes.

graeme_p

10:41 am on Sep 14, 2018 (gmt 0)

I wonder what they are going to do with this?

Leosghost

11:16 am on Sep 14, 2018 (gmt 0)

I wonder what they are going to do with this?

Try to find a way to slap ads around it whilst using it to keep the visitor in their walled garden ?

engine

11:48 am on Sep 14, 2018 (gmt 0)

I wonder what they are going to do with this?

From the horses mouth...

Understanding the text that appears on images is important for improving experiences, such as a more relevant photo search or the incorporation of text into screen readers that make Facebook more accessible for the visually impaired. Understanding text in images along with the context in which it appears also helps our systems proactively identify inappropriate or harmful content and keep our community safe.

Leosghost

12:11 pm on Sep 14, 2018 (gmt 0)

So..audio ads for the blind..and the amelioration of the "no breasts here" facebook algo..