Metadata-Version: 2.4
Name: pixgs
Version: 0.0.1a0
Summary: PixGS: Pixel-Space Diffusion for Direct 3D Gaussian Splat Generation
Home-page: https://github.com/jaco-bro/pixgs
Author: J Joe
Author-email: backupjjoe@gmail.com
License: Apache-2.0
Requires-Python: >=3.11
Description-Content-Type: text/markdown
Requires-Dist: torch
Dynamic: author
Dynamic: author-email
Dynamic: description
Dynamic: description-content-type
Dynamic: home-page
Dynamic: license
Dynamic: requires-dist
Dynamic: requires-python
Dynamic: summary

# PixGS

A minimal PyTorch implementation of [PixGS: Pixel-Space Diffusion for Direct 3D Gaussian Splat Generation](https://arxiv.org/abs/2607.01803v2), which is a single-stage pixel-space diffusion model that directly generates 3D Gaussian Splats from text or image conditioning. This repository is a compact, self-contained reimplementation intended for experimentation and understanding of the core ideas from the paper.

## Features

* Flow Matching objective
* Multi-view DiT backbone
* Pixel-aligned Gaussian attribute representation
* Plücker ray conditioning
* Differentiable Gaussian splat renderer
  * Pure PyTorch renderer
  * Optional `gsplat` backend
* Multi-stage training schedule
* RGB, depth, normal, LPIPS, and LoG supervision
* Toy synthetic dataset for end-to-end testing

## Included

* End-to-end training pipeline
* Sampling
* PLY export
* Smoke test
* Procedural toy dataset (`ToyBlobs`)

## Running

Smoke test:

```bash
python pixgs.py --smoke
```

Train on the toy dataset:

```bash
python pixgs.py
```

Use the CUDA `gsplat` renderer:

```bash
python pixgs.py --renderer gsplat
```

Outputs are written to `out_pixgs_mini/`.

## Reference

**PixGS: Pixel-Space Diffusion for Direct 3D Gaussian Splat Generation**

Duy Cao, Phong Nguyen-Ha

https://arxiv.org/abs/2607.01803v2
