3DFroMLLM: 3D Prototype Generation only from Pretrained Multimodal LLMs
By: Noor Ahmed , Cameron Braunstein , Steffen Eger and more
Potential Business Impact:
Makes computers build 3D shapes from words.
Recent Multi-Modal Large Language Models (MLLMs) have demonstrated strong capabilities in learning joint representations from text and images. However, their spatial reasoning remains limited. We introduce 3DFroMLLM, a novel framework that enables the generation of 3D object prototypes directly from MLLMs, including geometry and part labels. Our pipeline is agentic, comprising a designer, coder, and visual inspector operating in a refinement loop. Notably, our approach requires no additional training data or detailed user instructions. Building on prior work in 2D generation, we demonstrate that rendered images produced by our framework can be effectively used for image classification pretraining tasks and outperforms previous methods by 15%. As a compelling real-world use case, we show that the generated prototypes can be leveraged to improve fine-grained vision-language models by using the rendered, part-labeled prototypes to fine-tune CLIP for part segmentation and achieving a 55% accuracy improvement without relying on any additional human-labeled data.
Similar Papers
Part-X-MLLM: Part-aware 3D Multimodal Large Language Model
CV and Pattern Recognition
Lets computers build and change 3D objects with words.
MLLM-For3D: Adapting Multimodal Large Language Model for 3D Reasoning Segmentation
CV and Pattern Recognition
Helps computers understand 3D spaces like humans.
MMPart: Harnessing Multi-Modal Large Language Models for Part-Aware 3D Generation
CV and Pattern Recognition
Builds 3D objects from pictures, showing their parts.