# IDEA-Research/DWPose

"Effective Whole-body Pose Estimation with Two-stages Distillation" (ICCV 2023, CV4Metaverse Workshop)

Repository: https://github.com/IDEA-Research/DWPose
Canonical: https://ross.abutalabs.com/products/dwpose
Language: Python
License: Apache-2.0
License Family: permissive
Topics: controlnet, knowledge-distillation, pose-estimation, stable-diffusion-webui
Last push: 2023-12-12T02:54:23+00:00

## Health v2 (maintenance only)
Score: 28/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 0, release rhythm 35, longevity 81
- inputs: {"age_days": 1134, "days_push": 995, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 2807, forks 176 (observed 2026-08-28T04:07:22.871769+00:00)

## What it is
DWPose is the official implementation of 'Effective Whole-body Pose Estimation with Two-stages Distillation' (ICCV 2023), providing whole-body human pose estimation models from tiny to large sizes. It can replace Openpose as a ControlNet preprocessor in Stable Diffusion workflows and is built on MMPose.

## Use cases
- estimate whole-body human pose from images
- use pose estimation as ControlNet preprocessor for Stable Diffusion image generation
- run pose estimation with ONNX models without installing mmcv
- animate characters with pose-guided image-to-video synthesis
- choose pose models of different sizes for speed vs accuracy tradeoffs

## When to choose
- you need accurate whole-body (body, face, hands) pose estimation
- you want a better pose preprocessor than Openpose for ControlNet
- you need lightweight ONNX pose models for inference

## When to avoid
- you need 3D pose estimation
- you need pose tracking across video frames rather than per-frame estimation
- you need a non-Python or non-GPU environment without ONNX support

## Facets
- artifact type: library
- maturity: maintenance
- function: computer-vision, machine-learning, image-processing, stable-diffusion
- domain: computer-vision, image-processing, machine-learning, artificial-intelligence
- platform: python, cross-platform
- tags: pose-estimation, whole-body-pose, knowledge-distillation, controlnet, onnx, mmpose, stable-diffusion-webui, human-pose, gpu

## Member repositories
- IDEA-Research/DWPose (main) score 28

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:07:22.871769+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T08:14:26.973161+00:00, confidence not recorded.
  - readme: https://github.com/IDEA-Research/DWPose (fetched 2026-08-28T04:07:22.871769+00:00, sha cf3f647344f7)
- Data as of 2026-08-30T08:39:29.467469+00:00.