News from the AI & ML world
DeeperML
@avalovelace1.github.io
//
A research team from Carnegie Mellon University has developed LegoGPT, an AI model that generates Lego brick designs from text prompts. This innovative tool allows users to input a text description and receive a physically stable design that can be built using Lego bricks. The team trained the AI on a dataset of over 47,000 Lego structures encompassing more than 28,000 unique 3D objects, including furniture, vehicles, and other items. The system is now available for free to the public on GitHub, unlocking new possibilities for Lego enthusiasts and AI researchers alike.
The way LegoGPT works is by first converting the text prompt into a ShapeNetCore mesh, which is then voxelized onto a 20 x 20 x 20 grid. "Legolization" determines the brick layout. The system then augments each shape with multiple structural variations by randomizing the brick layout while preserving the overall shape. To ensure stability, the AI performs a stability analysis on each variation, removing any designs that might fall apart. This process ensures that the final output is not only visually appealing but also physically sound.
While the current version of LegoGPT is limited to designs within a 20 x 20 x 20 grid using eight basic brick types, the researchers envision a future where personalized product creation is easily accessible. Researcher Ava Pun told The Register, "We're dreaming of a future where making stuff becomes super personal! Imagine you just type what you want or show us a picture of a chair, and boom – we could actually make that product and ship it to you in just a week or two." The team emphasizes that LegoGPT's designs can be assembled both manually by humans and automatically by robotic arms, making it a versatile tool for creators of all skill levels.
ImgSrc: sm.pcmag.com
References :
The way LegoGPT works is by first converting the text prompt into a ShapeNetCore mesh, which is then voxelized onto a 20 x 20 x 20 grid. "Legolization" determines the brick layout. The system then augments each shape with multiple structural variations by randomizing the brick layout while preserving the overall shape. To ensure stability, the AI performs a stability analysis on each variation, removing any designs that might fall apart. This process ensures that the final output is not only visually appealing but also physically sound.
While the current version of LegoGPT is limited to designs within a 20 x 20 x 20 grid using eight basic brick types, the researchers envision a future where personalized product creation is easily accessible. Researcher Ava Pun told The Register, "We're dreaming of a future where making stuff becomes super personal! Imagine you just type what you want or show us a picture of a chair, and boom – we could actually make that product and ship it to you in just a week or two." The team emphasizes that LegoGPT's designs can be assembled both manually by humans and automatically by robotic arms, making it a versatile tool for creators of all skill levels.
ImgSrc: sm.pcmag.com
Share:
References :
- The Register - Software: At last, an AI model we can really get behind: LegoGPT takes a text prompt and spits out a physically stable design.…
- PCMag Middle East ai: LegoGPT eliminates AI Weirdness, Creates Brick Designs You Can Actually Build
- www.tomshardware.com: LegoGPT creates Lego designs using AI and text inputs — tool now available for free to the public
Classification:
- HashTags: #LegoGPT #AI #CarnegieMellon
- Company: Carnegie Mellon University
- Target: LEGO Enthusiasts
- Attacker: Carnegie Mellon University
- Product: LegoGPT
- Feature: Text-to-Lego
- Malware: LegoGPT
- Type: AI
- Severity: Informative