Skip to content

Image/video upscaling GUI app based on BRSGAN & RealSR_JPEG

License

Notifications You must be signed in to change notification settings

mtemelkov/QualityScaler

 
 

Repository files navigation

QualityScaler

Image/video upscaling & enhancement GUI app based on BRSGAN & RealSR_JPEG

Immagine 2022-04-03 164456

Credits.

BSRGAN - https://github.com/cszn/BSRGAN | https://arxiv.org/abs/2103.14006

RealSR_JPEG - https://github.com/jixiaozhong/RealSR | https://arxiv.org/pdf/2005.01996.pdf

Installation.

QualityScaler is completely portable; just download, unzip and execute the file .exe

Supported AI Backend.

  • Nvidia Cuda [v10.2]
  • CPU [works without GPU, but is very slow]

Features.

  • Easy to use GUI
  • Images and video upscale
  • Drag&drop files [image/multiple images/video]
  • Different upscale factors:
    • auto - automatic choose best upscale factor for the GPU used (to avoid running out of VRAM)
    • x1 - mantain same resolution but reconstruct & enhance the image (ideal for bigger images)
    • x2 - upscale factor 2: 500x500px -> 1000x1000px
    • x4 - upscale factor 4: 500x500px -> 2000x2000px
  • Cpu and Gpu [cuda] backend
  • Compatible images - PNG, JPEG, BMP, WEBP, TIF
  • Compatible video - MP4, WEBM, GIF, MKV, FLV, AVI, MOV

Next steps.

  • Use both model for the upscale
  • Include audio for upscaled video
  • Support for other GPUs (AMD, Intel) with new backend

Example.

Original photo (200x200 px)

220px-Lenna_test_image

BSRGAN (800x800 px)

220px-Lenna_test_image_resized_BSRGAN_x0

RealSR_JPEG (800x800 px)

220px-Lenna_test_image_resized_RealSR_JPEG_x0

About

Image/video upscaling GUI app based on BRSGAN & RealSR_JPEG

Resources

License

Stars

Watchers

Forks

Packages

No packages published

Languages

  • Python 100.0%