OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
AI Models

Synergynet

Synergynet is an open-source research model and code repository for estimating 3D facial geometry from images by combining a 3D morphable face model with facial landmarks. Its documented outputs include 3D landmarks, a face mesh and head-pose information.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Synergynet — OpenViST project profile in AI Models.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use Synergynet when a local research or development workflow needs image-based 3D facial geometry outputs and can accommodate the repository's data, dependencies and setup steps.

Who it is for

  • Computer-vision researchers exploring 3D facial alignment and reconstruction
  • Developers integrating facial geometry estimates into image-processing projects

Documented uses

  • Extract 3D facial landmarks and a face mesh from an image.
  • Estimate face pose as yaw, pitch and roll with translation.
  • Run the provided single-image inference or benchmark workflow.
  • Train the model using the documented training data and script.
Before you choose

Requirements & limits

The code repository is public and identified as MIT licensed. The documented workflow is a local setup with downloaded data and pretrained weights.

  • The simplified API returns 68 3D landmarks, 53,215 mesh points and face-pose values.
  • The default inference path requires a compatible GPU, though the README describes changes for CPU use.
  • The setup requires external model data and pretrained weights, as well as compiling Sim3DR and FaceBoxes.
  • The repository documents Python 3.8 and PyTorch 1.9 as its built/tested environment.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Prepare Synergynet for single-image inference

Set up the repository, dependencies, model data and pretrained weights required by the documented inference workflow.

Before you start

  • Python 3.8 and a compatible GPU for the default inference path
  • Conda
  • PyTorch 1.9, Torchvision, OpenCV, SciPy, Matplotlib and Cython

Steps

  1. Clone the Synergynet repository and create and activate a Conda environment.

  2. Install the listed prerequisite packages.

  3. Download the two required data packages and extract them under the repository root.

  4. Download pretrained weights and place the model under the repository's pretrained/ directory.

  5. Compile Sim3DR and FaceBoxes using the build scripts documented in the README.

Expected outcome

The repository has the documented dependencies, data, weights and compiled components for the next inference stage.

Guide limitations

  • The documented default inference requires a compatible GPU.
  • The README describes CPU inference as requiring edits to CUDA calls and loading pretrained weights on CPU.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
14 October 2026
Original website title
choyingw/SynergyNet · Official homepage title · retained in its source language
How we review sourcesSuggest a correction