How to install Deep Dream on Windows 10 (Nvidia Cuda supported GPU)

Note: If your graphics card isn't CUDA supported, things might not work.

1. Download Python

To download Python, open Command Prompt and type in

python

2. Download Anaconda

Download here https://www.anaconda.com/products/individual

3. Open Anaconda (Powershell Prompt) and run the following lines of code

conda create -n tf_gpu_env python=3.9   
conda activate tf_gpu_env
conda install tensorflow-gpu -c anaconda
conda install cudnn -c conda-forge 
conda install cudatoolkit -c anaconda
conda install -c conda-forge opencv ffmpeg
conda install -c conda-forge matplotlib
conda install -c anaconda pillow
conda install -c conda-forge glob2

4. Activate / Check if Tensorflow is running on your gpu

python

‎

import tensorflow as tf
print('Num GPUs Available', len(tf.config.list_physical_devices('GPU')))
exit()

‎

‎

How to use these scripts

1. Open Anaconda

2. Navigate to the deep_dream folder by using

cd path/to/deep_dream

3. Activate the gpu environment

Note: You have to do this everytime you restart anaconda.

conda activate tf-gpu

4. Run one of the following scripts

- Convert image(s) to Deep Dream

1. Put all images in the images_to_deep_dream folder.
2. All converted deep dream images will be generated in the deep_dream_images_input folder.
3. All converted images can be found in the deep_dream_images_output folder.

‎

python dream_images.py

- Video (mp4) to deep dream video

1. Put video in the video_to_deep_dream folder.
2. Generated video can be found in the video_output folder.

Tip If you want to stop the process press ctrl+c twice. When the video hasn't finished generating yet and you start the script again, it will continue where it has left of.

python dream_video.py

- Generate video frames from picture

1. Put a picture in the dream_on_image_input folder.
2. Generated pictures can be found in the dream_on_images_output folder.

Tip If you want to stop the process press ctrl+c twice.

python dream_on.py

- Generate video from images frames

1. Generate a video from the images frames created from
   • dream_images.py
   • dream_on.py
2. Video can be found in the video_output folder.

‎

python dream_images_to_video.py