# Google "Deep Dream" image processor **URL:** \ **Category:** Miscellaneous and Personal Stuff I Must Share\ **Created:** [July 7, 2015, 8:50pm UTC](https://boards.straightdope.com/t/google-deep-dream-image-processor/724504 "2015-07-07T20:50:02Z")\ **Posts on this page:** 20\ **Page:** 1
**Author:** ![Machine\_Elf](https://avatars.discourse-cdn.com/v4/letter/m/82dd89/32.png) [@Machine\_Elf](https://boards.straightdope.com/u/Machine_Elf)\ **Post date:** [July 7, 2015, 8:50pm UTC](https://boards.straightdope.com/t/google-deep-dream-image-processor/724504/1 "2015-07-07T20:50:02Z")
Holy crap, this is some mind-melting stuff. Google’s “Deep Dream” software takes images, identifies certain features and builds on them, often arriving at results that are completely different from the source image. Some fine still examples [here.](http://gizmodo.com/googles-dream-robot-is-running-wild-across-the-internet-1715839224) The really trippy shit happens when it processes video. For example, [here’s an unmodified clip from Fear and Loathing in Las Vegas](https://www.youtube.com/watch?feature=player_detailpage&v=erCSJIt5BKo#t=46) And [here’s what that same clip looks like after Deep Dream has had its way with it.](https://www.youtube.com/watch?v=oyxSerkkP4o&feature=player_embedded) :eek: ---
**Author:** ![Sattua](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/sattua/32/11906_2.png) [@Sattua](https://boards.straightdope.com/u/Sattua)\ **Post date:** [July 7, 2015, 8:56pm UTC](https://boards.straightdope.com/t/google-deep-dream-image-processor/724504/2 "2015-07-07T20:56:23Z")
I saw the stills a couple of weeks ago. Video is interesting, too. ---
**Author:** ![Sunspace](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/sunspace/32/1250_2.png) [@Sunspace](https://boards.straightdope.com/u/Sunspace)\ **Post date:** [July 7, 2015, 11:43pm UTC](https://boards.straightdope.com/t/google-deep-dream-image-processor/724504/3 "2015-07-07T23:43:10Z")
This is the neural-network thing, isn’t it? You train it to recognise, say, a puppy, and then turn it loose on non-puppy input and see what it does… I’m convinced something like this explains people seeing faces in things. ---
**Author:** ![Snowboarder\_Bo](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/snowboarder_bo/32/229_2.png) [@Snowboarder\_Bo](https://boards.straightdope.com/u/Snowboarder_Bo)\ **Post date:** [July 7, 2015, 11:57pm UTC](https://boards.straightdope.com/t/google-deep-dream-image-processor/724504/4 "2015-07-07T23:57:31Z")
> [@Machine\_Elf](#): > > The really trippy shit happens when it processes video. For example, [here’s an unmodified clip from Fear and Loathing in Las Vegas](https://www.youtube.com/watch?feature=player_detailpage&v=erCSJIt5BKo#t=46) > > And [here’s what that same clip looks like after Deep Dream has had its way with it.](https://www.youtube.com/watch?v=oyxSerkkP4o&feature=player_embedded) :eek: Aye, that’s what a dog-centric acid trip might look like alright. I wonder what it would do for an already trippy movie; like what would _Koyaanisqatsi_ look like after Deep Dreaming it? Hmmmmm, I might have to try that out… ---
**Author:** ![Sattua](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/sattua/32/11906_2.png) [@Sattua](https://boards.straightdope.com/u/Sattua)\ **Post date:** [July 8, 2015, 1:18am UTC](https://boards.straightdope.com/t/google-deep-dream-image-processor/724504/5 "2015-07-08T01:18:10Z")
> [@Sunspace](#): > > I’m convinced something like this explains people seeing faces in things. Sure. What’s miraculous about the human brain is that we _aren_’t seeing things that aren’t there, all the time. ---
**Author:** ![Snowboarder\_Bo](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/snowboarder_bo/32/229_2.png) [@Snowboarder\_Bo](https://boards.straightdope.com/u/Snowboarder_Bo)\ **Post date:** [July 8, 2015, 2:17am UTC](https://boards.straightdope.com/t/google-deep-dream-image-processor/724504/6 "2015-07-08T02:17:45Z")
VICE just put up a story featuring porn pics run through Deep Dream… it’s pretty twisted, totally NSFW and contains at least one image that could disrupt your own deep dreams… forever [What Computers Dream of When They Look at Porn (NSFW)](http://motherboard.vice.com/read/what-computers-dream-of-when-they-look-at-porn-nsfw) Note: Edited to hide the actual link. Some images are NSFW- ITD ---
**Author:** ![Machine\_Elf](https://avatars.discourse-cdn.com/v4/letter/m/82dd89/32.png) [@Machine\_Elf](https://boards.straightdope.com/u/Machine_Elf)\ **Post date:** [July 8, 2015, 3:01am UTC](https://boards.straightdope.com/t/google-deep-dream-image-processor/724504/7 "2015-07-08T03:01:26Z")
> [@Snowboarder\_Bo](#): > > [What Computers Dream of When They Look at Porn (NSFW)](http://motherboard.vice.com/read/what-computers-dream-of-when-they-look-at-porn-nsfw) Not sure if you’re talking about the face-full-of-dicks image, but that one made me LMAO. ---
**Author:** ![Fuzzy\_Dunlop](https://avatars.discourse-cdn.com/v4/letter/f/848f3c/32.png) [@Fuzzy\_Dunlop](https://boards.straightdope.com/u/Fuzzy_Dunlop)\ **Post date:** [July 8, 2015, 2:58pm UTC](https://boards.straightdope.com/t/google-deep-dream-image-processor/724504/8 "2015-07-08T14:58:30Z")
Of course it would be a **Machine Elf** that comes to show us freaky trippy images. ---
**Author:** ![G0sp3l](https://avatars.discourse-cdn.com/v4/letter/g/c68b51/32.png) [@G0sp3l](https://boards.straightdope.com/u/G0sp3l)\ **Post date:** [July 8, 2015, 8:51pm UTC](https://boards.straightdope.com/t/google-deep-dream-image-processor/724504/9 "2015-07-08T20:51:44Z")
Is there a mislabeled Nightmare Fuel switch somewhere? That donut almost made me sick. ---
**Author:** ![Sunspace](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/sunspace/32/1250_2.png) [@Sunspace](https://boards.straightdope.com/u/Sunspace)\ **Post date:** [July 8, 2015, 9:18pm UTC](https://boards.straightdope.com/t/google-deep-dream-image-processor/724504/10 "2015-07-08T21:18:51Z")
I downloaded the software last night and I’m going to try setting it up on my computer. I want to see what happens when I turn it loose on my sketchbooks. Edit: [Link.](http://googleresearch.blogspot.ca/2015/07/deepdream-code-example-for-visualizing.html) ---
**Author:** ![AHunter3](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/ahunter3/32/368_2.png) [@AHunter3](https://boards.straightdope.com/u/AHunter3)\ **Post date:** [July 8, 2015, 9:33pm UTC](https://boards.straightdope.com/t/google-deep-dream-image-processor/724504/11 "2015-07-08T21:33:33Z")
Did this algorithm get frontloaded with the notion that it should see/find animal faces? I mean, it isn’t drawing flowers or skyscrapers, it’s inventing faces, most of them of the earthly but nonhuman design. ---
**Author:** ![Skywatcher](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/skywatcher/32/254_2.png) [@Skywatcher](https://boards.straightdope.com/u/Skywatcher)\ **Post date:** [July 9, 2015, 5:35am UTC](https://boards.straightdope.com/t/google-deep-dream-image-processor/724504/12 "2015-07-09T05:35:08Z")
> [@Sunspace](#): > > I downloaded the software last night and I’m going to try setting it up on my computer. I want to see what happens when I turn it loose on my sketchbooks. I downloaded it and can’t figure out how to work the thing. Wanted to see what it does with [this](https://vimeo.com/88065833). ---
**Author:** ![Senegoid](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/senegoid/32/6606_2.png) [@Senegoid](https://boards.straightdope.com/u/Senegoid)\ **Post date:** [July 9, 2015, 6:28am UTC](https://boards.straightdope.com/t/google-deep-dream-image-processor/724504/13 "2015-07-09T06:28:29Z")
This guy gave it photos of the Republican candidates to chew on. [Google’s AI had nightmares about the GOP presidential candidates. This is what it saw.](http://www.washingtonpost.com/blogs/govbeat/wp/2015/07/08/googles-ai-had-nightmares-about-the-gop-presidential-candidates-this-is-what-it-saw/?tid=pm_politics_pop_b) Scroll down for the pictures. ---
**Author:** ![Senegoid](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/senegoid/32/6606_2.png) [@Senegoid](https://boards.straightdope.com/u/Senegoid)\ **Post date:** [July 9, 2015, 3:22pm UTC](https://boards.straightdope.com/t/google-deep-dream-image-processor/724504/14 "2015-07-09T15:22:48Z")
These pictures look like a team effort by Salvador Dali, Picasso, and Hieronymous Bosch. ---
**Author:** ![Machine\_Elf](https://avatars.discourse-cdn.com/v4/letter/m/82dd89/32.png) [@Machine\_Elf](https://boards.straightdope.com/u/Machine_Elf)\ **Post date:** [July 9, 2015, 4:34pm UTC](https://boards.straightdope.com/t/google-deep-dream-image-processor/724504/15 "2015-07-09T16:34:22Z")
> [@Senegoid](#): > > These pictures look like a team effort by Salvador Dali, Picasso, and Hieronymous Bosch. Add H.R. Giger to that list, and I’d say you’ve got it. > [@Senegoid](#): > > This guy gave it photos of the Republican candidates to chew on. > > Google’s AI had nightmares about the GOP presidential candidates. This is what it saw. Scroll down for the pictures. Dang, Bobby Jindal is cute as a button! The Puppyslug Nebula was great, too. Reminded me of the [three wolves/moon T-shirt.](http://knowyourmeme.com/memes/three-wolf-moon) The doghead pizza was a little unsettling, though. :eek: ---
**Author:** ![Telperion](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/telperion/32/509_2.png) [@Telperion](https://boards.straightdope.com/u/Telperion)\ **Post date:** [July 9, 2015, 8:45pm UTC](https://boards.straightdope.com/t/google-deep-dream-image-processor/724504/16 "2015-07-09T20:45:17Z")
I dread the day when we will have to explain to a sentient AI that _Grimm_ is, in fact, not a documentary. ---
**Author:** ![MindsEye\_Watering](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/mindseye_watering/32/3619_2.png) [@MindsEye\_Watering](https://boards.straightdope.com/u/MindsEye_Watering)\ **Post date:** [July 9, 2015, 10:27pm UTC](https://boards.straightdope.com/t/google-deep-dream-image-processor/724504/17 "2015-07-09T22:27:59Z")
Maybe I missed it upthread, but why are there so many “suggestions” of dog faces in the images? ---
**Author:** ![Senegoid](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/senegoid/32/6606_2.png) [@Senegoid](https://boards.straightdope.com/u/Senegoid)\ **Post date:** [July 10, 2015, 12:36am UTC](https://boards.straightdope.com/t/google-deep-dream-image-processor/724504/18 "2015-07-10T00:36:07Z")
> [@MindsEye\_Watering](#): > > Maybe I missed it upthread, but why are there so many “suggestions” of dog faces in the images? The “neural network” is “trained” by showing it a vastly large number of known pictures, and letting the AI software figure out for itself what their common salient features are. Apparently, the version that they release to the public had been “trained” with a large number of dog pictures, although there were other pictures in the training set too. That’s why you also see some birds, fish, and other strange stuff showing up. Somewhere on-line I saw some other pictures, produced by a network that had been trained mostly with pictures of buildings. Then, given a picture of a forest, it turned all the trees into pagodas. I’ll come back with a cite if I can find it again. ETA: Okay, here’s something for starters: [Why Google’s nightmare AI is putting demon puppies everywhere](http://www.washingtonpost.com/blogs/govbeat/wp/2015/07/08/why-googles-nightmare-ai-is-putting-demon-puppies-everywhere/) > [@](#): > > The network is called “BVLC GoogLeNet.” It’s based on a model designed by Google researchers that was one of the top performers at an international machine vision competition last year. Researchers at the Berkeley Vision and Learning Center made their own version and trained it on the 1.2 million images listed here. > > There are pictures of butterflies, sea slugs, microphones, maracas, and hares — but none, it seems, of humans. Furthermore, there are an unusually large number of dog photos, because researchers wanted to teach the neural network how to distinguish between dog breeds. > > Imagine if, from infanthood, your parents only showed you pictures of dogs. Wouldn’t you go a little dog crazy? Wouldn’t you start seeing dogs everywhere, even behind closed eyes? ---
**Author:** ![Senegoid](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/senegoid/32/6606_2.png) [@Senegoid](https://boards.straightdope.com/u/Senegoid)\ **Post date:** [July 10, 2015, 1:01am UTC](https://boards.straightdope.com/t/google-deep-dream-image-processor/724504/19 "2015-07-10T01:01:40Z")
I’m still looking for the one I saw earlier that showed a whole forest of trees turned into pagodas. In the meantime, here are some other bits and pieces: [Inceptionism: Going Deeper into Neural Networks](http://googleresearch.blogspot.com/2015/06/inceptionism-going-deeper-into-neural.html), June 17, 2015. Blog post from Google Research discussing a bit about what is going on in training their networks. Shows one example of one tree turned into a pagoda. Also, lots of clouds turned into strange fish, and other examples. [What is #DeepDream and why is everyone getting so weird with it?](http://killscreendaily.com/articles/what-deepdream-everyone-getting-weird/), June 7, 2015, from Kill Screen Daily. Also some discussion of how it works and examples. [DeepDream reddit page.](https://www.reddit.com/r/deepdream/) Everyone is posting links to their pictures here, along with some discussions. [What are deepdream images? How do I make my own? Can I do audio/video? Why are there dogs everywhere?!](https://www.reddit.com/r/deepdream/comments/3cawxb/what_are_deepdream_images_how_do_i_make_my_own/) Discussion of the techniques, instructions how you can make your own, why there are so many dogs. ---
**Author:** ![Senegoid](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/senegoid/32/6606_2.png) [@Senegoid](https://boards.straightdope.com/u/Senegoid)\ **Post date:** [July 10, 2015, 1:14am UTC](https://boards.straightdope.com/t/google-deep-dream-image-processor/724504/20 "2015-07-10T01:14:49Z")
Here’s the one I was looking for. It was trained on landscapes. This doesn’t show what the source picture was, but the result was a bunch of pagodas: [https://twitter.com/Synapticism/status/618490397958868992/photo/1](https://twitter.com/Synapticism/status/618490397958868992/photo/1) [Next page](https://boards.straightdope.com/t/google-deep-dream-image-processor/724504.md?page=2)