Janspiry/Image-Super-Resolution-via-Iterative-Refinement

Health v2 · maintenance only

Flags: no_releases

Adoption not part of the score

3923 stars · 484 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-29, confidence not recorded

An unofficial PyTorch implementation of SR3 (Image Super-Resolution via Iterative Refinement), a diffusion-based model for image super-resolution. It supports conditional upscaling (e.g., 64x64 to 512x512), unconditional face generation, and includes training utilities with multi-GPU support and logging.

Use cases

• upscale low-resolution images with a diffusion model
• train an SR3 super-resolution model in PyTorch
• generate faces unconditionally with DDPM
• reproduce the SR3 paper results
• run pretrained super-resolution on Colab
• experiment with conditional diffusion models

When to choose

• you want a PyTorch implementation of SR3 to study or extend
• you need diffusion-based image upscaling with pretrained weights
• you want a codebase for experimenting with conditional DDPMs

When to avoid

• you need a production-ready or officially supported super-resolution tool
• you want the latest diffusion SR techniques rather than a 2021-era reproduction
• you lack a GPU, since training and inference are compute-heavy

Facets

library · maturity maintenance

machine-learning deep-learning image-processing stable-diffusion computer-vision image-processing deep-learning machine-learning python super-resolution diffusion-models ddpm sr3 pytorch image-generation unofficial-implementation gpu linux

Member repositories

For agents

markdown · JSON · MCP: product_card(name="Janspiry/Image-Super-Resolution-via-Iterative-Refinement")