Wednesday, January 11, 2023

Point-E in Rhino: Generate Objects with OpenAi-Technology by Textprompts



Point-E creates point clouds of objects based on text input. Resolution is still relatively low to keep usability high, but the results are already amazing.

Point-E is the latest open AI web service that OpenAI released, following amazing tools like Dall-E for generating images and ChatGPT for generating text-based answers for almost any text prompt.

It is now usable in Rhino through the free version of the Adapative Parts Environment plug-in.

Multiple objects can be generated simultaneously and point clouds are even colored.

Currently, Rhenso is working on an automated meshing algorithm for Release 7.

If you are already using Release 8 beta, you should try the new ShrinkWrap feature with the result of Point-E.

To use it follow these steps:
  • Install Adaptive Parts Environment (APE) through the package manager (win only) or by our installer
  • Download and open this Example (3dm and Grasshopper file)
  • Use the APE Panel and enter your query
  • Copy and change the model as often as you like

2 comments:

Kristoffer said...

That looks very very cool! I have played around with Shap-E on hugginface.com, but would love to try this out directly in Rhino.
Is there any chance of getting the example file (3dm and grasshopper file)? I can not access the link anymore..
Br.
Kristoffer

Carlos Pérez said...

Hi Kristoffer, please contact directly support@rhenso.com