---
task_categories:
- text-to-3d
- image-to-3d
- image-text-to-image
- any-to-any
tags:
- Camera
- 3D Vision
- Spatial AI
- Physical AI
- World Model
- Camera Parameter
- DeepFashion
- Generation

---

# DeepFashion-Camera

![camera map collage](analysis/DeepFashion-Camera.png)

Per-image camera parameter annotations for the **DeepFashion** dataset
(full-body fashion model photographs; 13,679 images), captioned by the [**Puffin-World**](https://github.com/KangLiao929/Puffin) model. More captioned datasets are provided in our [**Puffin-16M**](https://kangliao929.github.io/projects/puffin-16m/) website.

The collage above visualizes the camera maps on sample images — each
pair shows the **up field** (green arrows: the projected gravity-up direction)
and the **latitude field** (colored contours: angle above/below the horizon).

## Format

One `.tar` per shard (`deepfashion_0000.tar` … ), each containing one `.json`
per image whose name matches the source image stem. Only the main person
photograph (the `image` field) is captioned; auxiliary pose / cloth views are
excluded.

Each JSON holds the predicted monocular camera parameters:

| Field | Meaning | Unit |
|-------|---------|------|
| `roll`  | camera roll        | radians |
| `pitch` | camera pitch       | radians |
| `vfov`  | vertical field-of-view | radians |
| `k1`    | radial distortion coefficient | – |
| `parse_ok` | whether the model output parsed within valid ranges | bool |

Example:

```json
{"roll": 0.0123, "pitch": 0.0087, "vfov": 0.5691, "k1": 0.0000, "parse_ok": true}
```

## Camera Parameter Distributions

Histograms of the predicted roll / pitch / vertical-FoV over the whole dataset
(proportion of valid samples per 10° bin; `parse_ok=False` excluded).

![deepfashion camera stats](analysis/deepfashion_camera_stats.png)

| split | roll μ / med / σ | pitch μ / med / σ | FoV μ / med / σ |
|-------|------------------|-------------------|-----------------|
| all (13,675) | 0.1° / 0.3° / 3.5° | 1.1° / 0.4° / 3.6° | 32.6° / 32.1° / 3.4° |

- **Roll** and **pitch** are both tightly centered near 0° (level, front-on
  full-body shots).
- **FoV** is remarkably consistent (σ ≈ 3.4°, the tightest of our captioned sets)
  — studio fashion photography frames the model with a near-constant focal length.

If you'd like a dataset with a more diverse and uniform distribution of camera parameters, please refer to our [Puffin-4M](https://huggingface.co/datasets/KangLiao/Puffin-4M) and [Puffin-16M](https://huggingface.co/datasets/KangLiao/Puffin-16M) datasets.

### Dataset Download
You can download the entire dataset using the following command:
```bash
hf download KangLiao/DeepFashion-Camera --repo-type dataset
```

### Caption Pipeline
Beyond this captioned dataset, we also release **a complete captioning pipeline** for annotating camera parameters for arbitrary datasets, analyzing camera parameter distributions, and visualizing the corresponding camera maps. The pipeline is available in our [GitHub repository](https://github.com/KangLiao929/Puffin).


### Citation
If you find the captioned dataset useful for your research or applications, please cite our paper using the following BibTeX:

```bibtex
  @article{liao2025puffin,
    title={Thinking with Camera: A Unified Multimodal Model for Camera-Centric Understanding and Generation},
    author={Liao, Kang and Wu, Size and Wu, Zhonghua and Jin, Linyi and Wang, Chao and Wang, Yikai and Wang, Fei and Li, Wei and Loy, Chen Change},
    journal={arXiv preprint arXiv:2510.08673},
    year={2025}
  }
```