World-Initializer-image-to-panorama Open-source Panorama Generation Model - Convert Single-view Images to 360° Panoramas with One Click

World Initializer Image To Panorama

Developed by TaiMingLu

A panorama generation model based on FluxFillPipeline, capable of converting single-view images into 360° panoramic images

Image Generation #Single Image to Panorama #360° Image Generation #Visual Inpainting Technology

Downloads 175

Release Time : 4/23/2025

Model Overview

Transforms single-view images into 2048×1024 equirectangular projection panoramas using visual inpainting technology, supporting text-guided generation

Model Features

Single Image to Panorama

Automatically converts single-view images of any size into 360° panoramic views

Text-Guided Generation

Supports guiding the panorama generation process through text prompts

Automatic Cropping

Automatically centers and crops input images into squares to fit the processing pipeline

Model Capabilities

Image Inpainting

Panorama Generation

Text-Guided Image Generation

Use Cases

Virtual Scene Construction

Game Scene Generation

Generates complete 360° game environments from single concept art

2048×1024 resolution equirectangular projection panorama

Virtual Tourism

Expands ordinary landscape photos into immersive panoramic views

Navigable panoramic environment

Creative Design

Concept Art Expansion

Creates complete environmental panoramas based on partial concept art

Stylistically consistent panoramic artwork

🚀 GenEx-World-Initializer 🧭🌍

GenEx World Initializer is a panorama generation pipeline built on the FluxFillPipeline. It can transform a single view image into a 360° panoramic image using vision-conditioned inpainting.

Key Information

Property	Details
Library Name	diffusers
Pipeline Tag	image-to-image
Inference Parameters	guidance_scale: 3.5
Example Input	example_input.jpg - GenEx Panoramic World Initialization (Panoramic generation from image crop)
Datasets	TaiMingLu/GenEx-DB-Panorama-World
Base Model	black-forest-labs/FLUX.1-Fill-dev
License	cc-by-4.0

🚀 Quick Start

This pipeline transforms a single view image into a 360° panoramic image. Here are the input and output details:

🖼️ Input: One image (any size, will be center-cropped to square)
🧠 Prompt: Optional text to guide panoramic generation
🎯 Output: 2048 × 1024 equirectangular image
🧩 Mask: Uses a fixed panoramic mask

Example Output

📦 Installation

Ensure you have the following dependencies installed:

diffusers>=0.33.1
transformers
numpy
pillow
sentencepiece

💻 Usage Examples

Basic Usage

from diffusers import DiffusionPipeline
from PIL import Image
import torch

pipe = DiffusionPipeline.from_pretrained(
    "genex-world/World-Initializer-image-to-panorama",
    custom_pipeline="genex_world_initializer_pipeline",  
    torch_dtype=torch.bfloat16,
    trust_remote_code=True
).to("cuda")

# Load your image (any resolution)
image = Image.open("example_input.jpg")

# Run inference
front_view, output = pipe(image=image)
output.images[0]

🏁 Mask

The following mask is used to train the inpainting diffuser and is used for inference automatically. Panoramic Mask

📄 License

This project is licensed under the cc-by-4.0 license.

✨ BibTex

@misc{lu2025genexgeneratingexplorableworld,
      title={GenEx: Generating an Explorable World}, 
      author={Taiming Lu and Tianmin Shu and Junfei Xiao and Luoxin Ye and Jiahao Wang and Cheng Peng and Chen Wei and Daniel Khashabi and Rama Chellappa and Alan Yuille and Jieneng Chen},
      year={2025},
      eprint={2412.09624},
      archivePrefix={arXiv},
      primaryClass={cs.CV},
      url={https://arxiv.org/abs/2412.09624}, 
}

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