Space ArtAI

https://sites.google.com/view/space-artai/inicio

https://www.youtube.com/watch?v=qGrjl30tOAA

We train a neural network with images of that artist. After this process we will have an artificial intelligence that through natural language can be asked to generate an image of a star, thus obtaining a result as if it were an order to the artist himself.

https://nasasearch.nasa.gov/search/images

It is a challenge for us to participate in a project with such a tight deadline. It helps us to learn how to present and sell a project with future potential. On the other hand, meeting other people and sharing knowledge with them is always enriching.

We have participated to test ourselves and get out of our comfort zone. We chose the art challenge because the result is very nice and it is a trending topic in the artificial intelligence scenario.

Luckily we are a disciplined team and we were able to distribute the tasks well, we had a clear idea of what we wanted and that helped us to have speed. The big challenges are always the speech and the marketing of the application. We have been helped by people we know for this task. We want to make a special mention to Julia that we met by chance in the presentation of the hackaton in madrid and we have based on her work to make this project, besides being a great person.

Our project has been based on a recently released artificial intelligence technology to generate image from natural language called "Stable diffusion" (https://stability.ai/blog/stable-diffusion-public-release).

We have developed a small server in python that contains this network and a multi-device web application in react (Facebook technology).

We started from the art of Julia (https://isisjul9.wixsite.com/julia-art/nebulosas) an artist we met. Also images from space and nasa.

https://colab.research.google.com/github/huggingface/notebooks/blob/main/diffusers/sd_dreambooth_training.ipynb

#SpaceArt #StableDiffusion #DeepLearning #GenerativeArt