How does Google filter out NSFW images in a GIS search?

[obbn]

@Rigamarole

It probably should be in the graphical editor, but you can type out the tags. Put the text between a (spoiler) and (/spoiler) tag, but replace the parentheses with brackets.

Got it, thanks for the help!

[BowlOfDucks]

Google reads the text and data surrounding images to get an idea of the context. This can filter out many pornographic images.

Then if you notice at the bottom of the image search page, there’s a “report offensive images” link. If you see an offensive image, you click that link and report the image. Which goes to help google filter out unwanted images.

[Colibri]

@obbn

Colibri, thanks I wasn’t aware of the rule.

For future reference, here’s the rules on NSFW images.

[kombatminipig]

@Swords_to_Plowshares

You can use machine learning to train computers on images just like you could with text.

Machine learning algorithms work like this for binary classification (ex. spam or not spam, porn or not porn):

Show the computer a lot of training examples, some with positive classification and some as negative. Tell it which ones are positive and which are negative.

From the training examples, it makes a guess and then learns from its mistakes if it guessed wrong. It figures out the similarities between the positive examples and the similarities between the negative examples. Keep doing this until it gets all the training examples right.

Then, unleash your algorithm on to the test set - in this case real world data. Monitor it and see how it does.

Combine this with a “report image” feature for real-time on-the-fly learning and you have a powerful algorithm that can automatically detect porn, just like a spam filter can detect spam.

You can, but with far less success. Text is still a tiny dataset in comparison to an image. The number of permutations in text is basically limited by the number of words in the dictionary, and further limited by the grammatically comprehensible ways of combining them. Even this only works to a very limited level, CleverBot being an example.
To be able to tell a human form from a random skin colored blotch demands association and context, two abilities that computers lack (computerized image recognition is built on statistics and comparison). This is the whole basis of captchas: the fact that the human mind can trivially recognize a number even though it’s upside down and warped while a computer can’t.

[BigT]

@kombatminipig

You can, but with far less success. Text is still a tiny dataset in comparison to an image. The number of permutations in text is basically limited by the number of words in the dictionary, and further limited by the grammatically comprehensible ways of combining them. Even this only works to a very limited level, CleverBot being an example.
To be able to tell a human form from a random skin colored blotch demands association and context, two abilities that computers lack (computerized image recognition is built on statistics and comparison). This is the whole basis of captchas: the fact that the human mind can trivially recognize a number even though it’s upside down and warped while a computer can’t.

But, as pointed out, all the system uses is text. There are very few images out there with no text describing them somewhere. All you have to do is find it.

[kombatminipig]

@BigT

But, as pointed out, all the system uses is text. There are very few images out there with no text describing them somewhere. All you have to do is find it.

Indeed. I was simply refuting the idea that there was any type of image analysis going on.