When it comes to 3D printing, sourcing files to print is half the fun or a daunting challenge, depending on who you ask. But if you want to take matters into your own hands, it can be worth investigating how to create 3D objects yourself. The learning curve for this sort of endeavor can be intimidating and takes a lot of time to master. Thankfully, we have geniuses like the team behind PartCrafter who are making the job of creating 3D objects from scratch easier than ever.
PartCrafter is an open source tool that you can find over at GitHub, built by individuals from Peking University, ByteDance (the same one that owns Tiktok), and Carnegie Mellon University. The software is designed to create 3D objects from scratch using 2D images as the input. It uses AI to evaluate the 2D image and constructs a mesh from the data.
According to the team, PartCrafter was trained on a database of around 50,000 3D objects that are made of about 300,000 individual pieces. One thing that PartCrafter seems to excel at is creating objects with a specific number of parts. You can scale up or down the number of components the 3D object is made from, and PartCrafter will come up with a way to create the object using the number you specify.
You may like
- LegoGPT creates Lego designs using AI and text inputs — tool now available for free to the public
- AI can track 3D printed parts back to specific machine that made them
- 6 Best Online Slicers to Try — 3D Slice on the Cloud
The team provided some example photos, which give us a good look at its ability to turn images into objects. Not only did they test this with individual objects, but also pictures of rooms. PartCrafter was able to reconstruct the layout of a room by accurately recreating the furniture in the picture and duplicating its position.