# cleardusk/3DDFA

The PyTorch improved version of TPAMI 2017 paper: Face Alignment in Full Pose Range: A 3D Total Solution.

Repository: https://github.com/cleardusk/3DDFA
Canonical: https://ross.abutalabs.com/products/3ddfa
Language: Python
License: MIT
License Family: permissive
Topics: face-alignment, pytorch, python, computer-vision, deep-learning, 3d-face, 3d, 3dmm
Last push: 2022-05-14T12:12:21+00:00

## Health v2 (maintenance only)
Score: 23/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 0, release rhythm 8, longevity 100
- inputs: {"age_days": 2987, "days_push": 1572, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 3677, forks 642 (observed 2026-08-28T04:08:13.674829+00:00)

## What it is
A PyTorch implementation of the TPAMI 2017 paper 'Face Alignment in Full Pose Range: A 3D Total Solution' (3DDFA). It fits a 3D Morphable Model to a single face image to produce 3D face shapes, 68 landmarks, pose estimation, depth images, PNCC/PAF features, and OBJ mesh serialization.

## Use cases
- fit a 3D face model to a photo
- detect 68 facial landmarks in large pose images
- estimate head pose from a single image
- generate depth maps of faces
- export 3D face meshes as OBJ files
- compute PNCC and PAF features for face analysis

## When to choose
- you need 3D face alignment or dense face reconstruction in Python with PyTorch
- you want landmark detection robust to large pose ranges
- you need face pose estimation, depth, or mesh outputs from a single image

## When to avoid
- you need the faster, more accurate successor (use 3DDFA_V2 instead)
- you need a maintained production pipeline - the repo is in maintenance mode with its last release in 2022
- you work outside Python/PyTorch and don't want to use the simple C++ port

## Facets
- artifact type: library
- maturity: maintenance
- function: computer-vision, machine-learning, deep-learning, image-processing
- domain: computer-vision, deep-learning
- platform: python, cross-platform
- tags: face-alignment, 3dmm, pytorch, landmarks, pose-estimation, 3d-face-reconstruction, 3d-face

## Member repositories
- cleardusk/3DDFA (main) score 23

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:08:13.674829+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 2026-08-29T18:31:15.473759+00:00, confidence not recorded.
  - readme: https://github.com/cleardusk/3DDFA (fetched 2026-08-28T04:08:13.674829+00:00, sha d9fa5332a12d)
- Data as of 2026-08-30T08:39:29.467469+00:00.