ChatGPT maker introduces Shap-E to create 3D models
We know OpenAI from ChatGPT and Dall-E, but it now has a new artificial intelligence tool: Shap-E. It is not a fitness tool, but a way to generate a 3D model based on an instruction through text.
Shap-E
In Dall-E you can generate works of art and photorealistic images by writing texts. The same goes for Shap-E, only you don’t get art from this, but 3D models. You can even provide these models with all kinds of textures. Because NeRF, which stands for neural radiance fields, is used, images can be generated that are complex and of high quality.
NeRF is a visual field synthesis technique widely used in the development of maps, navigation, virtual reality, augmented reality and robotics. Shap-E can generate implicit function parameters, allowing you to create complex 3D models. Think of a chair that looks like an avocado, or a banana plane, as Gizmochina has come up with very creatively.
NeRF
In the past, neural radiation fields were not used, but point clouds. In this, points were distributed that then fall into a 3D coordinate system and are spread over the subject in order to represent it that way. Each cell has a fixed size.
Unlike previous OpenAI tools, this is not a tool that you consult via the browser. You can go to Github for this and download the program from there to use locally on your PC. You can then generate very professional looking 3D models in seconds. You can imagine that it is a very attractive tool for making games and apps, but also for scientific research and product development.
Related articles
Artificial intelligence for 3D models
In any case, it is another step in the wonderful world of artificial intelligence. Shap-E again comes from a company that has already earned its spurs in AI, so this tool also promises to be a success. Moreover, we can imagine that this is also a less fraud-prone tool, or could you also use it to make the pope in a puffer jacket?
Are you going to use Shap-E and if so, what for? Leave it now in the comments.
