AI is making 3D models that you can 3D print
Science & Technology
Introduction
Introduction
Lately, I’ve noticed a trend of articles popping up all over the internet talking about AI 3D model generators, and I was immediately interested. I've known about 2D AI image generators for years; they’ve grown in capability and popularity but weren’t particularly useful to me unless you’re a graphic designer or artist. 3D models, on the other hand, have broader applications, such as in video games, AR, or VR worlds, and they can even be 3D printed.
Google's Stable Dream Fusion
Google recently developed a system named Stable Dream Fusion, based on the stable diffusion model that creates 2D images. Their system converts Simple Text prompts like "a frog wearing a sweater" or "a corgi nurse plush toy" into full-color 3D models. I couldn’t resist and went to their website to borrow the example 3D models available, clean them up for 3D printing, and share them on Printables. You're welcome to try printing them; the link is in the description.
Open Source Alternatives
Following Google’s lead, it seems someone created an open-source version that works similarly to Stable Dream Fusion. Browsing through GitHub, it appears that if you have some programming skills and an Nvidia GPU, you could run this open-source version on your computer to generate AI models from text prompts. While I lack the computer and programming skills to get it running, I am eager to hear about anyone else’s experience with it. Again, the link is in the description.
Current Limitations
It's important to note that this method isn’t perfect yet, which is understandable given that this is a new development. The GitHub for the open-source version mentions that many text prompts may fail, and there's a known issue called the Janus problem, where models with faces may generate multiple faces depending on the viewing angle, often with eerie results. Nonetheless, the technology is fascinating, and its potential improvements are highly anticipated.
Future Implications
Considering the technology’s rapid advancements, just a few years ago, 2D AI image generators were primitive and would've required powerful computers to run. Nowadays, you can find many free AI image generators running directly in your browser, integrated into tools like Blender or Canva, producing impressive results.
For 3D sculptors, this technology could be revolutionary. Imagine entering a text prompt to generate a 3D mesh close to your vision, then making modifications and final tweaks to significantly accelerate your workflow from idea to finished 3D model.
Future Potential
Years down the line, this technology could evolve to the point where someone could voice a request to a smart TV or speaker, like “Hey, I want a Victorian-inspired lamp.” AI 3D model generators could then preview this model, take user feedback for any changes, and forward it to local manufacturing or 3D printing facilities for production and delivery. Such a system could open up vast opportunities for unique, customized products, adding diversity and intrigue to the market.
Conclusion
Overall, AI-generated 3D models hold exciting possibilities for the future. If you end up printing any of the AI-generated models I shared, let me know your thoughts. Be kind to each other and happy printing!
Keywords
- AI 3D model generators
- Stable Dream Fusion
- Open-source 3D modeling
- 3D printing
- Text-to-3D models
- Future of 3D modeling
FAQs
Q1: What is Stable Dream Fusion?
A1: Stable Dream Fusion is a system developed by Google that converts text prompts into full-color 3D models.
Q2: Can I try generating 3D models using AI on my own computer?
A2: Yes, there is an open-source version available on GitHub that you can try if you have some programming skills and an Nvidia GPU.
Q3: Are there any limitations to AI-generated 3D models?
A3: Yes, the technology is not perfect yet. Many text prompts may fail, and models with faces may suffer from the Janus problem, producing multiple eerie faces from different angles.
Q4: What are the future implications of this technology?
A4: In the future, AI 3D model generators could allow users to voice requests for custom items, preview the models, make adjustments and send them for local manufacturing or 3D printing, ushering in an era of unique, personalized products.
Q5: How can this technology benefit 3D sculptors?
A5: 3D sculptors could use AI-generated meshes as a starting point, speeding up the process from idea to finished model by modifying and refining the auto-generated base models.