A Dockerized deepdream Cloud Deamon

Google recently released the deepdream software package which uses the Caffe Deep Learning Library and all of the code runs in an iPython notebook.

So let's make it brain-dead simple to launch your very own deepdreaming server (in the cloud, on an Ubuntu machine, Mac via Docker, and maybe even Windows if you try out Kitematic by Docker)!

Motivation

Because the installation procedure for Caffe is not trivial, I decided to create a self-contained Caffe/deepdream Docker image which has everything you need to generate your own deepdream art. In order to make the image very portable, it uses the CPU version of Caffe and comes bundled with the GoogLeNet model.

The compilation procedure was done on Docker Hub and the final image can be pulled down via

docker pull visionai/clouddream

The docker image is 2.5GB, but it contains a precompiled version of Caffe, all of the python dependencies, as well as the pretrained GoogLeNet model.

For those of you who are new to Docker, I hope you will pick up some valuable engineering skills and tips along the way. Docker makes it very easy to bundle complex software. If you're somebody like me who likes a clean Mac OS X on a personal laptop, and do the heavy-lifting in the cloud, then read on.

Instructions for use

We will be monitoring the inputs directory for source images and dumping results into the outputs directory. Additionally, there is a simple Python-based HTTP server running on port 80 which server the resulting images.

Prerequisite:

You've launched a Cloud instance using a VPS provider like DigitalOcean and this instance has Docker running. If you don't know about DigitalOcean, then you should give them a try. You can lauch a Docker-ready cloud instance in a few minutes. If you're going to set up a new DigitalOcean account, consider using my referral link: https://www.digitalocean.com/?refcode=64f90f652091.

Let's say our cloud instance is at the address 1.2.3.4

ssh root@1.2.3.4
git clone https://github.com/VISIONAI/clouddream.git
./start.sh

Then from your local machine you can just scp images into the inputs directory inside deepdream as follows:

#From your local machine
scp "images/*jpg" root@1.2.3.4:~/clouddream/deepdream/inputs/

You should now be able to visit http://1.2.3.4 in your browser and see the resulting images appear one by one.

Changing the default parameters

Inside deepdream.py you'll notice that I'm using

frame = deepdream(net, img, end='conv2/3x3')

But you can try different layers such as:

frame = deepdream(net, img, end='inception_3b/5x5_reduce')

Feeding deepdream your own images

You can prepopulate the inputs directory with a flat directory of jpeg images, or you can scp them from you local machine.

I applied this demo to all of the images in the PASCAL VOC 2011 dataset. You can find the resulting images on the deepdream.vision.ai server.

Credits

The included Dockerfile is an extended version of https://github.com/taras-sereda/docker_ubuntu_caffe

Which is a modification from the original Caffe CPU master Dockerfile tleyden: https://github.com/tleyden/docker/tree/master/caffe/cpu/master

This dockerfile uses the deepdream code from: https://github.com/google/deepdream

License

MIT License. Have fun. Never stop learning.

--Enjoy! The vision.ai team

Pentusha / clouddream