---
dataset_info:
  features:
  - name: image
    dtype: image
  - name: messages
    list:
    - name: content
      list:
      - name: index
        dtype: int64
      - name: text
        dtype: string
      - name: type
        dtype: string
    - name: role
      dtype: string
  splits:
  - name: train
    num_bytes: 2393634762.584
    num_examples: 6378
  download_size: 2975326101
  dataset_size: 2393634762.584
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
original_dataset_id: Voxel51/MPII_Human_Pose_Dataset
processing_description: Filtered missing body keypoint annotations, reformatted keypoint
  coordinates for Molmo
source_datasets:
- extended
tags:
- remyx
pretty_name: PoseText
size_categories:
- 1K<n<10K
---

![image/webp](https://cdn-uploads.huggingface.co/production/uploads/647777304ae93470ffc28913/kZIoNBVWOVhXnJVvDu1HH.webp)

# Dataset Card for the PoseText Dataset

The PoseText Dataset can be used to enhance vision-language model performance in the task of human pose estimation.

## Dataset Details

### Dataset Description

Parsing body keypoints from the [Voxel51/MPII_Human_Pose_Dataset](https://huggingface.co/datasets/Voxel51/MPII_Human_Pose_Dataset), parsing into a text-based format used in [Molmo](https://huggingface.co/allenai/Molmo-7B-O-0924/discussions/1)

```python
import re
import numpy as np

def extract_points(molmo_output, image_w, image_h):
    all_points = []
    for match in re.finditer(r'x\d*="\s*([0-9]+(?:\.[0-9]+)?)"\s+y\d*="\s*([0-9]+(?:\.[0-9]+)?)"', molmo_output):
        try:
            point = [float(match.group(i)) for i in range(1, 3)]
        except ValueError:
            pass
        else:
            point = np.array(point)
            if np.max(point) > 100:
                # Treat as an invalid output
                continue
            point /= 100.0
            point = point * np.array([image_w, image_h])
            all_points.append(point)
    return all_points
```

- **Curated by:** [remyx.ai]
- **Language(s) (NLP):** [en]

## Citation
If you found this resource useful, please consider citing:
```
@misc{posetext2024,
  title={PoseText},
  author={Terry Rodriguez and Salma Mayorquin},
  organization={Remyx AI},
  year={2024},
  month={September},
  note = {Dataset},
  url = {https://huggingface.co/datasets/salma-remyx/PoseText}
}
```