par

KO | EN

gitlite — search

#python #react #android #machine-learning #rust #linux #docker #macos #java #windows #kubernetes #awesome-list #openai #open-source

(★ 10)

[NeurIPS 2025] Conditional Panoramic Image Generation via Masked Autoregressive Modeling

par Latest Version Download

Download Latest Version (.zip)

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

# Conditional Panoramic Image Generation via Masked Autoregressive Modeling

[![arXiv](https://img.shields.io/badge/arXiv-2505.16862-b31b1b.svg)](https://arxiv.org/abs/2505.16862)
[![Project Website](https://img.shields.io/badge/🔗-Project_Website-blue.svg)](https://wang-chaoyang.github.io/project/par)

</div>

<div>
  <p align="center" style="font-size: larger;">
    <strong>NeurIPS 2025</strong>
  </p>
</div>

## Requirements

Install other pip packages via `pip install -r requirements.txt`.

## Preparation

Download the [pre-trained model](https://github.com/baaivision/NOVA/blob/main/docs/model_zoo.md) and put them under `data/pretrain`.

We use [Matterport3D](https://niessner.github.io/Matterport) in our experiments and follow [PanFusion](https://github.com/chengzhag/PanFusion) for data preparation. The data should be organized as follows.
```
data
├── Matterport3D
│   ├── mp3d_skybox
│   │   ├── 1LXtFkjw3qL
│   │       ├── matterport_skybox_images
│   │       └── matterport_stitched_images   
│   │   ├── train.npy   
│   │   └── test.npy  
│   └── caption
│       ├── 1LXtFkjw3qL_0b22fa63d0f54a529c525afbf2e8bb25.txt
└── pretrain
```

## Demo

Download the [checkpoint](https://huggingface.co/chaoyangw/par/blob/main/par_w1024.pt) and put it under `data/pretrain/ckpt/par_w1024.pt`.

```
accelerate launch scripts/demo.py --ptr data/pretrain/nova-d48w1024-sdxl1024 --ckpt data/pretrain/ckpt/par_w1024.pt --prompt=[Prompt]
```

## Train
The codes in [scripts](scripts/train.py) is launched by accelerate. The images and captions are specified by `path` and `textpath`, respectively.
```
accelerate launch scripts/train.py configs/cfg.yaml ptr=[ptr] path=[path] textpath=[textpath] env.o=[project path]
```

## Inference
```
accelerate launch scripts/train.py configs/cfg.yaml ptr=[ptr] path=[path] textpath=[textpath] env.o=[project path] eval_only=true 
```

## Citation

If you find our work interesting, please kindly consider citing our paper:

```bibtex
@article{wang2025conditional,
    title={Conditional Panoramic Image Generation via Masked Autoregressive Modeling},
    author={Wang, Chaoyang and Li, Xiangtai and Qi, Lu and Lin, Xiaofan and Bai, Jinbin and Zhou, Qianyu and Tong, Yunhai},
    journal={arXiv preprint arXiv:2505.16862},
    year={2025}
}
```

## License

Apache License 2.0

par

par Latest Version Download

Today's picks

Was this content helpful?

Recent Feedback

par

par Latest Version Download

Today's picks

Was this content helpful?

Recent Feedback

Download README