YMA
Available for software engineering opportunitiesCV
Project index
09 / 09AI / computer vision

AI Image Colorizer

A Streamlit tool that applies a pre-trained deep-learning colorization model to grayscale images.

GitHub (opens in a new tab)
Field
AI / computer vision
Editorial luminance-to-color pipeline created for AI Image Colorizer
Original editorial cover · not a product screenshot
Overview

System brief.

The application combines a lightweight web interface with OpenCV's DNN module and a Caffe colorization model based on the work of Zhang et al.

  • Upload or local-folder input
  • Side-by-side source and result comparison
  • Downloadable colorized output
  • Large model assets managed with Git LFS
Technical index

Stack.

  • 01Python
  • 02Streamlit
  • 03OpenCV DNN
  • 04Caffe
  • 05Git LFS
Inference / 01

A direct image-to-image workflow

A grayscale input is processed through OpenCV's DNN interface using the supplied pre-trained Caffe model. The Streamlit layer presents both versions together and exposes the generated result for download.

Model / 02

Research-backed, pre-trained colorization

The project credits the colorization method introduced by Zhang and collaborators at ECCV 2016. Model files are distributed separately through Git LFS because of their size.