### Setup Conda Environment

Source: https://github.com/zhaoyujie2002/langsplatv2/blob/main/README.md

Create and activate the project environment using the provided YAML file.

```shell
conda env create --file environment.yml
conda activate langsplat_v2
```

--------------------------------

### Clone the Repository

Source: https://github.com/zhaoyujie2002/langsplatv2/blob/main/README.md

Use these commands to clone the repository with submodules via SSH or HTTPS.

```shell
# SSH
git clone git@github.com:ZhaoYujie2002/LangSplatV2.git --recursive
```

```shell
# HTTPS
git clone https://github.com/ZhaoYujie2002/LangSplatV2.git --recursive
```

--------------------------------

### Preprocess scene data

Source: https://github.com/zhaoyujie2002/langsplatv2/blob/main/README.md

Generate language features for the scenes before training.

```shell
python preprocess.py --dataset_path $dataset_path
```

--------------------------------

### Evaluate datasets

Source: https://github.com/zhaoyujie2002/langsplatv2/blob/main/README.md

Evaluation scripts for different dataset types.

```shell
bash eval_lerf.sh $scene_name $model_idx $checkpoint
```

```shell
bash eval_3d_ovs.sh $scene_name $model_idx $checkpoint
```

```shell
bash eval_mip_nerf360.sh $scene_name $model_idx $checkpoint
```

--------------------------------

### Dataset Directory Structure

Source: https://github.com/zhaoyujie2002/langsplatv2/blob/main/README.md

Required directory layout for custom scenes, following the 3D Gaussian Splatting format.

```text
<dataset_name>
|---images
|   |---<image 0>
|   |---<image 1>
|   |---...
|---output
|   |---<dataset_name>
|   |   |---point_cloud/iteration_30000/point_cloud.ply
|   |   |---cameras.json
|   |   |---cfg_args
|   |   |---chkpnt30000.pth
|   |   |---input.ply
|---sparse
    |---0
        |---cameras.bin
        |---images.bin
        |---points3D.bin
```

--------------------------------

### Evaluate LERF dataset

Source: https://github.com/zhaoyujie2002/langsplatv2/blob/main/README.md

Run evaluation on the LERF dataset using the specified scene, model index, and checkpoint.

```shell
bash eval_lerf.sh $scene_name $model_idx $checkpoint
```

--------------------------------

### Train LangSplat V2

Source: https://github.com/zhaoyujie2002/langsplatv2/blob/main/README.md

Train the global semantic codebook and sparse coefficient field.

```shell
bash train.sh $dataset_root_path $scene_name $model_idx
```

--------------------------------

### BibTeX Citations for Research Series

Source: https://github.com/zhaoyujie2002/langsplatv2/blob/main/README.md

Citations for related LangSplat research papers.

```bibtex
@inproceedings{qin2024langsplat,
  title={Langsplat: 3d language gaussian splatting},
  author={Qin, Minghan and Li, Wanhua and Zhou, Jiawei and Wang, Haoqian and Pfister, Hanspeter},
  booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition},
  pages={20051--20060},
  year={2024}
}
```

```bibtex
@inproceedings{li20254d,
  title={4d langsplat: 4d language gaussian splatting via multimodal large language models},
  author={Li, Wanhua and Zhou, Renping and Zhou, Jiawei and Song, Yingwei and Herter, Johannes and Qin, Minghan and Huang, Gao and Pfister, Hanspeter},
  booktitle={Proceedings of the Computer Vision and Pattern Recognition Conference},
  pages={22001--22011},
  year={2025}
}
```

--------------------------------

### BibTeX Citation for LangSplat V2

Source: https://github.com/zhaoyujie2002/langsplatv2/blob/main/README.md

Citation entry for the LangSplat V2 paper.

```bibtex
@article{li2025langsplatv2,
      title={LangSplatV2: High-dimensional 3D Language Gaussian Splatting with 450+ FPS}, 
      author={Wanhua Li and Yujie Zhao and Minghan Qin and Yang Liu and Yuanhao Cai and Chuang Gan and Hanspeter Pfister},
      year={2025},
      journal={Advances in Neural Information Processing Systems},
      eprint={2507.07136},
      archivePrefix={arXiv},
      primaryClass={cs.GR},
      url={https://arxiv.org/abs/2507.07136}, 
}
```

=== COMPLETE CONTENT === This response contains all available snippets from this library. No additional content exists. Do not make further requests.