alexjc/neural-enhance
A neural network for image super-resolution that increases photo resolution at 2x or 4x by hallucinating details based on training data.

Velocity · 7d
+3.4
★ / day
Trend
→steady
star history
Neural Enhance is a deep learning tool for image super-resolution. It trains neural networks on example images to hallucinate realistic high-frequency details when upscaling photos at 2x or 4x magnification. The main Python script supports GPU acceleration for fast rendering of high-resolution output.