Datagen Documentation#

Datagen provides an all-purpose generator of synthetic images of human faces and bodies, providing perfectly annotated datapoints for training and testing your neural network.

🚀 Quickstart

Generate your first datapoint in ten minutes!

⚙️ API Reference

Learn more about the Datagen SDK

🎞 Modalities

Explore the data & metadata that you get with each datapoint

🐱 DGUtils

A GitHub repo of tools that help you process your data

📚 Use-Case Templates

Pre-made scripts to solve your Computer Vision task

Sample datapoints#

Here is a sampling of the data you can generate on our platform: visual spectrum images, semantic segmentation, depth maps, nir (near-infrared), and normal maps:



  • Overview: Installation and usage of our SDK, which you can use to generate data requests to order custom datapoints

  • About data requests: A detailed guide to the structure of a custom data request, through which you exercise granular control over the datapoints you order from our system

  • Playbooks: Guidelines for how best to use our platform for specific neural network tasks

  • About Datagen: Who we are and what we do