Reference

AI Photo Colorization Glossary

Clear definitions of the technologies behind AI photo colorization — from diffusion models and DeOldify to neural colorization and semantic segmentation. Built for family historians, archivists, and curious readers.

AI Diffusion Models

Generative AI models that gradually add detail to noise until a full color image emerges.

neural colorization generative adversarial network

DeOldify

An open-source deep learning project for AI colorization and restoration of historical images.

generative adversarial network neural colorization

Neural Colorization

Using deep neural networks to automatically add color to grayscale photographs.

ai diffusion models generative adversarial network

Generative Adversarial Network (GAN)

A neural network architecture where two models compete to produce realistic synthetic images.

deoldify neural colorization ai diffusion models

Photo Restoration

Repairing physical and digital damage in old photographs before or alongside colorization.

neural colorization semantic segmentation

Semantic Segmentation

AI technique that labels each pixel by object type — sky, skin, foliage — to guide colorization.

neural colorization photo restoration