# una-dinosauria/3d-pose-baseline

A simple baseline for 3d human pose estimation in tensorflow. Presented at ICCV 17.

Repository: https://github.com/una-dinosauria/3d-pose-baseline
Canonical: https://ross.abutalabs.com/products/3d-pose-baseline
Language: Python
License: MIT
License Family: permissive
Topics: tensorflow, computer-vision, 3d-vision, baseline, iccv-2017, iccv-17
Last push: 2020-09-26T02:27:50+00:00

## Health v2 (maintenance only)
Score: 32/100 (v2, computed 2026-09-03T02:39:23.370411+00:00)
- activity 0, release rhythm 35, longevity 100
- inputs: {"age_days": 3422, "days_push": 2168, "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 1459, forks 357 (observed 2026-08-28T04:04:47.186926+00:00)

## What it is
A TensorFlow implementation of a simple yet effective baseline for 3D human pose estimation from 2D keypoints, published at ICCV 2017. It includes training, evaluation, and visualization code on the Human3.6M dataset.

## Use cases
- estimate 3d human pose from 2d keypoints
- reproduce the iccv 2017 3d pose baseline results
- train a pose lifting model on human3.6m
- benchmark my 3d pose estimation model against a baseline
- visualize 3d human skeleton poses from video
- learn how 2d-to-3d pose lifting works

## When to choose
- you need a well-known, citable baseline for 2D-to-3D human pose lifting
- you are working with the Human3.6M dataset and ground-truth 2D detections
- you want compact, readable research code to build upon

## When to avoid
- you need state-of-the-art accuracy on modern benchmarks
- you require 2D detections from an external detector (SH detections are no longer supported)
- you need TensorFlow 2.x or actively maintained code

## Facets
- artifact type: library
- maturity: maintenance
- function: machine-learning, computer-vision, deep-learning
- domain: computer-vision, machine-learning, deep-learning
- platform: python
- tags: 3d-pose-estimation, human-pose, tensorflow, iccv-2017, research-code, human36m, gpu

## Member repositories
- una-dinosauria/3d-pose-baseline (main) score 32

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:47.186926+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 2026-08-30T04:35:28.641296+00:00, confidence not recorded.
  - readme: https://github.com/una-dinosauria/3d-pose-baseline (fetched 2026-08-28T04:04:47.186926+00:00, sha 9114643107b0)
- Data as of 2026-08-30T08:39:29.467469+00:00.