README.md

---
license: openrail
task_categories:
- image-to-image
- keypoint-detection
- pose-estimation
language:
- en
tags:
- controlnet
- openpose
size_categories:
- 10K<n<100K
dataset_info:
  features:
  - name: id
    dtype: int64
  - name: image
    dtype: image
  - name: conditioning_image
    dtype: image
  - name: text
    dtype: string
---

# Important Notice

This is a copy of raulc0399/open_pose_controlnet replacing [openpose](https://github.com/CMU-Perceptual-Computing-Lab/openpose) conditioning images 
with [DW pose](https://github.com/IDEA-Research/DWPose).

# Dataset for training controlnet models with conditioning images as Human Pose

the entries have been taken from this dataset
[ptx0/photo-concept-bucket](https://huggingface.co/datasets/ptx0/photo-concept-bucket)

the open pose images have been generated with
[controlnet_aux](https://github.com/huggingface/controlnet_aux)

for the scripts to download the files, generate the openpose and the dataset please refer to:
[raulc0399/dataset_scripts](https://github.com/raulc0399/dataset_scripts)

# Openpose to DWpose processing

then openpose images have been replaced using OpenCV_Onnx code available
[here](https://github.com/IDEA-Research/DWPose/blob/opencv_onnx/ControlNet-v1-1-nightly/annotator/dwpose/__init__.py).