# bearpaw/pytorch-pose

A PyTorch toolkit for 2D Human Pose Estimation.

Repository: https://github.com/bearpaw/pytorch-pose
Canonical: https://ross.abutalabs.com/products/pytorch-pose
Language: Python
License: GPL-3.0
License Family: copyleft
Topics: pose, human-pose-estimation, pytorch, mpii-dataset, mscoco-keypoint, hourglass-network, pose-estimation
Last push: 2021-12-20T20:15:47+00:00

## Health v2 (maintenance only)
Score: 32/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 0, release rhythm 35, longevity 100
- inputs: {"age_days": 3410, "days_push": 1717, "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 1121, forks 249 (observed 2026-08-28T04:03:39.908476+00:00)

## What it is
A PyTorch toolkit implementing a general pipeline for 2D single-human pose estimation, with training, inference, and evaluation interfaces plus data loaders for MPII, LSP, FLIC, and MSCOCO datasets. It includes pretrained models such as the stacked hourglass network and supports multi-GPU training and result visualization.

## Use cases
- estimate human body keypoints from images
- train a pose estimation model on MPII or COCO
- evaluate a stacked hourglass network on pose benchmarks
- run pose estimation with pretrained models
- augment pose datasets for training
- visualize keypoint predictions during training

## When to choose
- you need a ready-made PyTorch pipeline for 2D human pose estimation on standard benchmarks
- you want pretrained hourglass models for quick experimentation
- you need multi-GPU training with data augmentation for pose datasets

## When to avoid
- you need 3D pose estimation or multi-person pose estimation
- you require modern PyTorch versions or Python 3 support, which are untested
- you need an actively maintained production library - the last release was 2021

## Facets
- artifact type: library
- maturity: maintenance
- function: machine-learning, deep-learning, computer-vision, image-processing
- domain: computer-vision, deep-learning, machine-learning
- platform: python
- tags: pose-estimation, human-pose, keypoint-detection, hourglass-network, pytorch, mpii, mscoco, research-code, linux, gpu

## Member repositories
- bearpaw/pytorch-pose (main) score 32

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:03:39.908476+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 2026-08-30T06:40:59.517095+00:00, confidence not recorded.
  - readme: https://github.com/bearpaw/pytorch-pose (fetched 2026-08-28T04:03:39.908476+00:00, sha 291b0f1c38bb)
- Data as of 2026-08-30T08:39:29.467469+00:00.