Average Ratings 0 Ratings
Average Ratings 0 Ratings
Description
Recent advancements in the realm of text-to-image synthesis have emerged from diffusion models that have been trained on vast amounts of image-text pairs. To successfully transition this methodology to 3D synthesis, it would necessitate extensive datasets of labeled 3D assets alongside effective architectures for denoising 3D information, both of which are currently lacking. In this study, we address these challenges by leveraging a pre-existing 2D text-to-image diffusion model to achieve text-to-3D synthesis. We propose a novel loss function grounded in probability density distillation that allows a 2D diffusion model to serve as a guiding principle for the optimization of a parametric image generator. By implementing this loss in a DeepDream-inspired approach, we refine a randomly initialized 3D model, specifically a Neural Radiance Field (NeRF), through gradient descent to ensure its 2D renderings from various angles exhibit a minimized loss. Consequently, the 3D representation generated from the specified text can be observed from multiple perspectives, illuminated with various lighting conditions, or seamlessly integrated into diverse 3D settings. This innovative method opens new avenues for the application of 3D modeling in creative and commercial fields.
Description
Recent advancements in text-based 3D object generation have yielded encouraging outcomes; however, leading methods generally need several GPU hours to create a single sample, which is a stark contrast to the latest generative image models capable of producing samples within seconds or minutes. In this study, we present a different approach to generating 3D objects that enables the creation of models in just 1-2 minutes using a single GPU. Our technique initiates by generating a synthetic view through a text-to-image diffusion model, followed by the development of a 3D point cloud using a second diffusion model that relies on the generated image for conditioning. Although our approach does not yet match the top-tier quality of existing methods, it offers a significantly faster sampling process, making it a valuable alternative for specific applications. Furthermore, we provide access to our pre-trained point cloud diffusion models, along with the evaluation code and additional models, available at this https URL. This contribution aims to facilitate further exploration and development in the realm of efficient 3D object generation.
API Access
Has API
No
API Access
Has API
No
Integrations
No details available.
Integrations
No details available.
Pricing Details
No price information available.
Free Trial
No
Free Version
No
Pricing Details
No price information available.
Free Trial
No
Free Version
No
Deployment
Web-Based
Yes
On-Premises
No
iPhone App
No
iPad App
No
Android App
No
Windows
No
Mac
No
Linux
No
Chromebook
No
Deployment
Web-Based
Yes
On-Premises
No
iPhone App
No
iPad App
No
Android App
No
Windows
No
Mac
No
Linux
No
Chromebook
No
Customer Support
Business Hours
No
Live Rep (24/7)
No
Online Support
Yes
Customer Support
Business Hours
No
Live Rep (24/7)
No
Online Support
Yes
Types of Training
Training Docs
Yes
Webinars
No
Live Training (Online)
No
In Person
No
Types of Training
Training Docs
Yes
Webinars
No
Live Training (Online)
No
In Person
No
Vendor Details
Company Name
DreamFusion
Website
dreamfusion3d.github.io
Vendor Details
Company Name
OpenAI
Founded
2015
Country
United States
Website
openai.com/research/point-e