TripoSR Open-Source 3D Generation Model - Free Deployment, Rapidly Reconstruct 3D Models from a Single Image

Triposr

Developed by stabilityai

TripoSR is a fast feed-forward 3D generation model jointly developed by Stability AI and Tripo AI, specializing in rapid 3D reconstruction from a single image.

3D Vision Open Source License:MIT #Single-image 3D Reconstruction #Game Development Optimization #Fast Feed-forward

Downloads 19.25k

Release Time : 2/29/2024

Model Overview

TripoSR is a single-image feed-forward 3D reconstruction model based on the LRM architecture, achieving efficient 3D generation through data optimization and training enhancements, making it particularly suitable for game development and rapid prototyping.

Model Features

Fast Generation

Utilizes a feed-forward architecture to achieve second-level 3D model generation, significantly improving efficiency.

Game Development Optimization

Output resource formats are compatible with mainstream game engines and can be directly used in development workflows.

Enhanced Generalization

Trained with improved rendering methods, it can better handle real-world images.

Model Capabilities

Single-image 3D Reconstruction

Rapid Mesh Generation

Texture Map Generation

Use Cases

Game Development

Rapid Prop Modeling

Quickly generate 3D models of game props from concept art.

Reduces traditional modeling time by over 50%.

Virtual Reality

Scene Rapid Construction

Convert 2D environment images into 3D scenes.

Enables real-time scene prototyping.

🚀 TripoSR

TripoSR is a fast and feed - forward 3D generative model jointly developed by Stability AI and Tripo AI, enabling single - image 3D reconstruction.

🚀 Quick Start

Try our new model: SF3D with several improvements such as faster generation and more game - ready assets.

The model is available here and we also have a demo.

Visual TripoSR is a fast and feed - forward 3D generative model developed in collaboration between Stability AI and Tripo AI.

📚 Documentation

🔍 Model Details

📋 Model Description

We closely follow LRM network architecture for the model design. TripoSR incorporates a series of technical advancements over the LRM model in terms of both data curation as well as model and training improvements. For more technical details and evaluations, please refer to our tech report.

Property	Details
Developed by	Stability AI, Tripo AI
Model Type	Feed - forward 3D reconstruction from a single image
License	MIT
Hardware	We train `TripoSR` for 5 days on 22 GPU nodes each with 8 A100 40GB GPUs

🌐 Model Sources

Repository: https://github.com/VAST - AI - Research/TripoSR
Tech report: https://arxiv.org/abs/2403.02151
Demo: https://huggingface.co/spaces/stabilityai/TripoSR

📊 Training Dataset

We use renders from the [Objaverse](https://objaverse.allenai.org/objaverse - 1.0) dataset, utilizing our enhanced rendering method that more closely replicate the distribution of images found in the real world, significantly improving our model’s ability to generalize. We selected a carefully curated subset of the Objaverse dataset for the training data, which is available under the CC - BY license.

💻 Usage

For usage instructions, please refer to our [TripoSR GitHub repository](https://github.com/VAST - AI - Research/TripoSR)
You can also try it in our gradio demo

⚠️ Important Note

The model should not be used to intentionally create or disseminate 3D models that people would foreseeably find disturbing, distressing, or offensive; or content that propagates historical or current stereotypes.

Featured Recommended AI Models

Empowering the Future, Your AI Solution Knowledge Base

English 简体中文繁體中文にほんご