OGIex

OGIex are scripts from building a docker image to running a demo of paper implementations by three or fewer commands and works.

For example, you can get a prediction result of stable-dreamsfusion by two commands:

# stage 1: build a docker image and start a docker container.
bash docker_buildrun.sh
# stage 2: predict a 3D object from a text (a hamburger), and then output a result into output dir (outputs/hamburger).
bash text_to_novelview.sh "a hamburger" outputs/hamburger

How to use

Git clone this repository:

git clone https://github.com/Obarads/OGIex.git

Select a method that you want to run from the Script dir list section, and then move current dir to the method dir. For example, if you want to run stable-dreamsfusion, you can run the following command:
```
cd OGIex/scripts/stable-dreamfusion
```

Run the commands in a README.md of the method dir. For example, you can run a demo of stable-dreamsfusion by two commands:

# stage 1: build a docker image and start a docker container.
bash docker_buildrun.sh

# stage 2: predict a 3D object from a text (a hamburger), and then output a result into output dir (outputs/hamburger).
bash text_to_novelview.sh "a hamburger" outputs/hamburger

Script dir list

The list of available scripts is below. Dirname, Code, and Paper columns denote script directory links, implementation links, and article links, respectively.

Dirname	Code	Paper
1~0
4dfy	Official	4D-fy: Text-to-4D Generation Using Hybrid Score Distillation Sampling
C・c
CoDeF	Official	CoDeF: Content Deformation Fields for Temporally Consistent Video Processing
D・c
dreamgaussian	Official	DreamGaussian: Generative Gaussian Splatting for Efficient 3D Content Creation
G・g
gaussian-splatting	Official	3D Gaussian Splatting for Real-Time Radiance Field Rendering
L・l
LucidDreamer	Official	LucidDreamer: Towards High-Fidelity Text-to-3D Generation via Interval Score Matching
M・m
Make-It-3D	Official	Make-It-3D: High-Fidelity 3D Creation from A Single Image with Diffusion Prior
MM-Diffusion	Official	MM-Diffusion: Learning Multi-Modal Diffusion Models for Joint Audio and Video Generation
O・o
One-2-3-45	Ofiicial	One-2-3-45: Any Single Image to 3D Mesh in 45 Seconds without Per-Shape Optimization
P・p
Point-Bind_Point-LLM	Official	Point-Bind & Point-LLM: Aligning Point Cloud with Multi-modality for 3D Understanding, Generation, and Instruction Following
point-e	Official	Point-E: A System for Generating 3D Point Clouds from Complex Prompts
R・r
realfusion	Official	RealFusion: 360° Reconstruction of Any Object from a Single Image
S・s
shap-e	Official	Shap-E: Generating Conditional 3D Implicit Functions
stable-dreamfusion	Unofficial	DreamFusion: Text-to-3D using 2D Diffusion
T・t
text2room	Official	Text2Room: Extracting Textured 3D Meshes from 2D Text-to-Image Models
threestudio	Official	threestudio
W・w
Wonder3D	Official	Wonder3D: Single Image to 3D using Cross-Domain Diffusion
Z・z
zero123	Official	Zero-1-to-3: Zero-shot One Image to 3D Object
zero123plus	Official	Zero123++: A Single Image to Consistent Multi-view Diffusion Base Model