UTrice

Paper

This repository is the official implementations of UTrice: Unifying Primitives in Differentiable Ray Tracing and Rasterization via Triangles for Particle-Based 3D Scenes.

We propose a differentiable triangle-based ray tracing pipeline that treats triangles as primitives without proxy geometry, achieving higher rendering quality while maintaining real-time performance. It also directly renders triangles optimized by Triangle Splatting, unifying the primitives used in novel-view-synthesis.

Requirements

• This project has been built and tested on Ubuntu 22.04 with Python 3.11 and CUDA 12.6.
• To reproduce our results, we recommend a GPU with 24GB VRAM and Ray Tracing (RT) Cores. We tested on an RTX 4500 Ada and RTX 6000 Ada.

Installation

Clone this repository with submodules:

git clone --recursive https://github.com/waizui/UTrice.git
cd UTrice

We use Conda to manage the environments. Please install Conda in advance, then execute the following commands to create and activate an environment.

conda env create -f environment.yaml
conda activate utrice

Then install these two modules to finish the installation.

pip install submodules/simple-knn --no-build-isolation
pip install submodules/diff-triangle-tracer --no-build-isolation

Training

To train our model, use the following command:

python train.py -s <path_to_scenes> -m <output_model_path> --eval

Add --outdoor to train on outdoor scenes:

python train.py -s <path_to_scenes> -m <output_model_path> --outdoor --eval

Rendering

To render a scene, use:

python render.py -m <path_to_model>

Evaluation

To evaluate the model, use:

python metrics.py -m <path_to_model>

Reproduce our results

We provide a convinent script to reproduce the results of our paper, to run this scripts, please put all scenes inside a sibiling directory of this project and rename it to datasets. It should look like:

<parent>/
 ├─ datasets/
 │   ├─ bicycle/
 │   ├─ garden/
 │   └─ ...
 └─ <UTrice>/

Then, run following command:

python eval_all.py

The results will be saved in the UTrice/output directory.

Datasets

The MipNeRF360 scenes are hosted by the paper authors here. You can find SfM datasets for Tanks&Temples and Deep Blending here.

Render Effects

We provide examples of ray tracing effects using --render_mode: dof, reflect, refract, envlight.

python render.py -m <path_to_model> --render_mode <render_mode>

Citation

@article{liu2025utrice,
  title = {UTrice: Unifying Primitives in Differentiable Ray Tracing and Rasterization via Triangles for Particle-Based 3D Scenes},
  author = {Changhe Liu and Ehsan Javanmardi and Naren Bao and Alex Orsholits and Manabu Tsukada},
  journal = {arXiv},
  year = {2025},
}

@misc{liu2025utriceunifyingprimitivesdifferentiable,
      title={UTrice: Unifying Primitives in Differentiable Ray Tracing and Rasterization via Triangles for Particle-Based 3D Scenes}, 
      author={Changhe Liu and Ehsan Javanmardi and Naren Bao and Alex Orsholits and Manabu Tsukada},
      year={2025},
      eprint={2512.04421},
      archivePrefix={arXiv},
      primaryClass={cs.CV},
      url={https://arxiv.org/abs/2512.04421}, 
}