Inspiration
We really liked the idea of existing products, such as Sisyphus which implement relaxing dynamic art with design elements.
What it does
We are taking this idea and stepping up a notch. Using artificial intelligence we can get different results with each run and then display these patterns dynamically, creating a similar relaxing effect but tailored to specific needs.
How we built it
We utilized information from the web and translated them into prompts for DALLE-3 image generation model. We used the generated image to create gcode commands that our robot turns into dynamic art.
Challenges we ran into
Our first challenge was choosing the right hardware solutions and prompt engineering the DALLE-3 output into images that we can turn into art. The collaboration between our hardware and software team was crucial.
Accomplishments that we're proud of
We managed to create a working prototype for our installation and successfully test the whole generative process - from world news to a dynamic canvas.
What we learned
Rapid prototyping and debugging. We expanded our knowledge of openAIs API usage. Converting image data into machine-readable commands for computerized interpretation and drawing.
What's next for ARTIGEN
All that is really left here is to just build the end product. We have a working prototype that is a proof of concept. We need to work more on prompt engineering and work on implementing the end idea to be more than just our first attempt at a prototype or a 3d render.
Log in or sign up for Devpost to join the conversation.