AI tools, scored for your job
Get paid to train AI
🤝 You're in the Free & Open Source library — community-built tools you can use without paying. A different concept from our scored AI directory.
Home › Free & Open Source › Writing › PanFusion
✦ Free & open sourceWriting

PanFusion

PyTorch implementation for text-to-360-degree panorama image generation using Stable Diffusion.

Text to 360° Panorama Generation with Stable Diffusion

CVPR 2024 highlight implementing PanFusion for immersive panorama creation and evaluation.

Free forever · open source · MIT
✓ Needs IT✓ Browser-based✓ Real-time collaboration
5.4
ZEKAI
Advanced 360° panorama AI
Free & open source · Writing
Ease of setup3
Usefulness8
Maintenance5.5
Community5
★ 273
GITHUB
Dec 2025
UPDATED
MIT
LICENSE
🛠 Don't want to self-host? We can set it up →
⚡ Quick answer

PanFusion is a PyTorch-based framework that tames Stable Diffusion for text-to-360-degree panorama image generation. It suits researchers and developers working on generative AI, 3D environments, and spatial imaging.

Runs on
Cross-platform
Hosting
Self-host
Languages
Python
License
MIT

What it is

The tool provides complete data preprocessing, training, testing, and evaluation code for generating 360-degree panoramas from text prompts. It includes modules for stitching Matterport3D skybox images, captioning with BLIP, rendering layouts, and training evaluation models like FAED and HorizonNet.

Who it's for

It suits machine learning researchers and computer vision engineers with access to multi-GPU hardware like NVIDIA A100s. It does not suit casual users or those without deep learning infrastructure and Python experience.

Key features

Stable Diffusion Panorama

Tames diffusion models for generating 360-degree equirectangular panorama images.

Layout-Conditioned Generation

Finetunes a ControlNet model for layout-conditioned panorama synthesis.

Wandb Logging Integration

Uses Weights & Biases for logging, visualizing, and identifying training issues.

FAED and HorizonNet

Includes training code for evaluation models to measure panorama and layout quality.

How to start

Up and running in 3 steps

1
Clone the repository and create the environment using conda env create -f environment.yaml
2
Download pretrained checkpoints and place them in the designated logs or weights folders
3
Run prediction scripts using main.py with the appropriate data and model arguments
Real talk

Our honest verdict

What's great

  • Complete pipeline covering preprocessing, training, testing, and evaluation
  • Supports both text-to-image generation and layout-conditioned panorama creation
  • Provides pretrained checkpoints for quick testing and inference
  • Includes scripts for stitching Matterport3D skybox images and generating captions

Where it falls short

  • Requires high-end hardware, specifically multiple NVIDIA A100 GPUs for training
  • Complex setup involving Anaconda, custom datasets, and external Matlab preprocessing tools
  • Heavy reliance on external dataset downloads like Matterport3D and specific tar archives
Cost

Pricing & license

Free under its MIT license, with no hosted cloud options mentioned in the README. The only cost is running the hardware required for training and inference.

Paid work

Open remote AI-training roles

877 open remote AI-training roles, updated November 16, 2026; most pay $20–$175 an hour. Zekai lists them free; the application goes to the employer’s platform.

See all open roles →
Compare

Alternatives

🎁 Free & open-source alternatives
Get more

Use it, or have us set it up

📄 Free PDF: PanFusion for Writing
A plain-language guide to getting your first result. Sent to your inbox.
🔔 64 people want a full video tutorial — add your vote
🛠 Don't want to set it up yourself?
We'll install, host & configure PanFusion for you — and train your team.
Get it done for you ↗
Questions

FAQ

What hardware is required to train PanFusion?

Training the text-to-image generation model requires 4x NVIDIA A100 GPUs, while evaluation models like FAED or HorizonNet can run on a single NVIDIA A100.

Which datasets are supported for training?

The framework primarily uses the Matterport3D skybox dataset and Matterport3DLayoutAnnotation dataset.

How do I evaluate the generated panoramas?

You can train and use the included FAED model or HorizonNet checkpoints to evaluate image and layout quality.

Radar

Latest AI news

✉️ The weekly free toolkit
The best new free & open-source tools for Writing — one short email a week, in your language.
Free · 1 email/week · unsubscribe anytime
Get paid to train AI →

See Zekai first in Google