LegoGPT is a groundbreaking approach that generates LEGO structures from user prompts while guaranteeing physical stability and buildability. By utilizing a large-scale dataset of LEGO designs and employing advanced techniques like physics-aware rollback during inference, the model creates aesthetically pleasing and diverse designs. The project also involves a text-based LEGO texturing method and offers a new dataset, StableText2Lego, with over 47,000 LEGO structures. These advancements highlight LegoGPT's potential for manual assembly as well as automated construction using robotic arms.
We introduce LegoGPT, the first approach for generating physically stable LEGO brick models from text prompts, ensuring that creations are both buildable and aesthetically appealing.
Our experiments show that LegoGPT produces stable, diverse, and aesthetically pleasing LEGO designs that align closely with the input text prompts, showcasing the model's effectiveness.
Collection
[
|
...
]