xverse-engine

XScene-UEPlugin

# XScene-UEPlugin <a href="./LICENSE"> <img alt="License" src="https://img.shields.io/badge/License-Apache_2.0-blue.svg"></a> English | [中文](./README_CN.md) [ <img src="UEPlugin/Media/image/XVERSE.jpg" width="600" /> ](http://xverse.cn/) --- ## Table of Contents - [Introduction](#introduction) - [Project Structure](#project-structure) - [Training Module](#training-module) - [UE Plugin](#ue-plugin) - [Getting Started](#getting-started) - [Roadmap](#roadmap) - [Contributing](#contributing) - [Release Note](#release-note) --- ## Introduction XScene-UEPlugin is an Unreal Engine 5 (UE5) plugin developed by XVERSE Technology Inc. (Shenzhen, China). It provides real-time visualization, management, editing, and scalable hybrid rendering of Gaussian Splatting models—a novel technique for reconstructing 3D scenes from multi-view photos. For more details, see [3D Gaussian Splatting](https://repo-sam.inria.fr/fungraph/3d-gaussian-splatting/). --- ## Project Structure ```text 📦 XScene-UEPlugin ├─ 📁 Training │ ├─ gaussian-splatting/ # Core algorithm and training scripts │ ├─ README.md # English training guide │ └─ README_CN.md # 中文训练指南 ├─ 📁 UEPlugin │ ├─ Config/ # Plugin configuration presets │ ├─ Content/ # Example assets and Niagara emitters │ ├─ Media/ # Documentation images and videos │ ├─ Plugin/ # Plugin source code │ ├─ README.md # English plugin guide │ └─ README_CN.md # 中文插件指南 ├─ LICENSE └─ README.md # Main overview file ``` --- ## Training Module The `Training` folder contains everything you need to train your own 3D Gaussian Splatting models from videos or image sequences. **Quick Links:** - [Training Guide (EN)](./Training/README.md) - [训练指南 (中文)](./Training/README_CN.md) ### Highlights - **Data Preparation:** Tools to convert multi-view images or videos into training-ready formats - **Model Configuration:** Easily tweak `.yaml` config files for custom resolution, point count, and learning rate - **Training Scripts:** Single-command launch on Windows using `XV3DTools.exe` or Python scripts for advanced users - **Monitoring:** Integrated TensorBoard support for loss curves, PSNR, and other metrics --- ## UE Plugin The `UEPlugin` folder contains all UE5 plugin assets, source code, and documentation. **Quick Links:** - [Plugin Guide (EN)](./UEPlugin/README.md) - [插件指南 (中文)](./UEPlugin/README_CN.md) ### Features - **Real-time Gaussian Splatting rendering using Niagara** - **Drag-and-drop .gspl assets into UE5 Content Browser** - **Fully Blueprint-compatible and extendable** - **Dynamic lighting and LOD generation** - **Hybrid rendering with native UE assets** - **Crop regions & pure VFX Niagara emitters** --- ## Getting Started ### 1. Clone the Repository ```bash git clone https://github.com/xverse-engine/XScene-UEPlugin.git cd XScene-UEPlugin ``` ### 2. Install Dependencies #### For Training Ensure you have Python 3.8+, PyTorch, and the required packages: ```bash pip install -r Training/gaussian-splatting/requirements.txt ``` #### For UE Plugin Install Unreal Engine 5.0+ and enable the **Niagara Plugin**. ### 3. Build & Run - **Training:** See `Training/README.md` or `README_CN.md` for training steps - **UE Plugin:** Copy `UEPlugin/Plugin` to your UE project `Plugins/` directory and enable `XVERSE3DGS` in the Plugin Browser --- ## Roadmap - [ ] Dynamic LOD Rendering in Editor & Runtime - [ ] Automatic Collision & Physics Proxy Generation - [ ] Real-time Preview Window for quick iterations - [ ] Interactive 3D-GS Asset Editing - [ ] Compression & Streaming of 3D-GS Assets Contributions and feature requests are welcome! --- ## Contributing 1. Fork the repository 2. Create a branch (`git checkout -b feature/my-feature`) 3. Commit your changes (`git commit -m "Add feature"`) 4. Push (`git push origin feature/my-feature`) 5. Submit a Pull Request 🎉 --- ## Release Note [v1.1.6](https://github.com/xverse-engine/XScene-UEPlugin/tree/v1.1.6) - Provides a convenient and simple training code and script - The directory structure has been adjusted to separate training and UE plugins [v1.1.5](https://github.com/xverse-engine/XScene-UEPlugin/tree/v1.1.5) - Support user-defined settings of training parameters - Modify 3DGS densification strategy to focus more on important areas - Support post-processing strategies for object reconstruction, which can effectively remove floaters - Support Gaussian models with spherical harmonic coefficients up to the 3rd degree <img src="UEPlugin/Media/image/compare/1.5-1.png" width="500" /> <img src="UEPlugin/Media/image/compare/1.5-2.png" width="500" /> [v1.1.4](https://github.com/xverse-engine/XScene-UEPlugin/tree/v1.1.4) - Support UE5.4 [v1.1.3](https://github.com/xverse-engine/XScene-UEPlugin/tree/v1.1.3) - Support dragging multiple ply files into the content directory - Fix the crash issue when the buffer asset position is incorrect - Fix bug in XV3DTools while using wrong path [v1.1.2](https://github.com/xverse-engine/XScene-UEPlugin/tree/v1.1.2) - Supports more types of ply header - Fixed flickers when moving objects - Update UI [v1.1.1](https://github.com/xverse-engine/XScene-UEPlugin/tree/v1.1.1) - New Model Clipping allowing the creation of a clean Niagara for VFX. - XV3DTools v1.1.1: Users can now adjust the training iteration. [v1.1.0](https://github.com/xverse-engine/XScene-UEPlugin/tree/v1.1.0) - Training tools in Windows system (XV3DTools v1.0) to train a given mp4 video to GaussianSplatting ply file - Automatic lod generation, supporting over 200,000 point clouds in niagara - fix bug in transform [v1.0.1](https://github.com/xverse-engine/XScene-UEPlugin/tree/v1.0.0) - Fix bug in v1.0.0：Error while packaging for Windows in UE5.2 and UE5.3 [v1.0.0](https://github.com/xverse-engine/XScene-UEPlugin/tree/v1.0.0) - Niagara-Based High-quality real-time visualizing and rendering for 3D Gaussian Splatting - Easily importing and converting from the original Gaussian Splatting scene (.ply file) to ours - Hybrid rendering with other UE assets - RTS and Geometry Editing - Apply VFX effect to Gaussian Splatting scene - Support dynamic illumination of Gaussian Splatting scene - Automatic enhancement of Gaussian Splatting scene ## Contributors <a href="https://github.com/xverse-engine/XV3DGS-UEPlugin/graphs/contributors"> <img src="https://contrib.rocks/image?repo=xverse-engine/XV3DGS-UEPlugin" /> </a>