Collections

Restore images

These models restore and improve images by fixing defects like blur, noise, and low resolution. Key capabilities:

  • Deblurring - Sharpen blurry images by reversing blur effects. Useful for old photos.
  • Denoising - Remove grain and artifacts by learning noise patterns.
  • Colorization - Add realistic color to black and white photos.
  • Face restoration - Improve the image quality of faces in old photos, or unrealistic AI generated faces.

Our Picks

Best photo restoration model: jingyunliang/swinir.

If you need to sharpen a blurry photo, upscale a small picture, or remove noise or compression artifacts, start with jingyunliang/swinir. It’s a fast and powerful model for many types of photo restoration. Another popular choice is megvii-research/nafnet.

Best colorization model: piddnad/ddcolor

The best model for adding color to black and white photos is piddnad/ddcolor. It runs faster and produces more vibrant results than other models. If you are looking for more diverse outputs, try cjwbw/bigcolor which will give you several options from one input.

Best face restoration model: tencentarc/gfpgan

If you’re looking for a face restoration model, start with tencentarc/gfpgan, which runs in a fraction of a second. If you need even more realistic faces, try sczhou/codeformer.

For more options, make sure to check out our image upscaling collection →

Recommended models