Skip to content

Latest commit

 

History

History
51 lines (29 loc) · 2.21 KB

README_macos.md

File metadata and controls

51 lines (29 loc) · 2.21 KB

RealESRGAN

We have provided the following models:

  1. realesr-animevideov3 (default)
  2. realesrgan-x4plus
  3. realesrgan-x4plus-anime

Command:

You may need to change file mode before running.

chmod u+x realesrgan-ncnn-vulkan

  1. ./realesrgan-ncnn-vulkan -i input.jpg -o output.png
  2. ./realesrgan-ncnn-vulkan -i input.jpg -o output.png -n realesr-animevideov3
  3. ./realesrgan-ncnn-vulkan -i input_folder -o outputfolder -n realesr-animevideov3 -s 2 -f jpg
  4. ./realesrgan-ncnn-vulkan -i input_folder -o outputfolder -n realesr-animevideov3 -s 4 -f jpg

Commands for enhancing anime videos:

  1. Use ffmpeg to extract frames from a video (Remember to create the folder tmp_frames ahead)

    ffmpeg -i onepiece_demo.mp4 -qscale:v 1 -qmin 1 -qmax 1 -vsync 0 tmp_frames/frame%08d.jpg

  2. Inference with Real-ESRGAN executable file (Remember to create the folder out_frames ahead)

    ./realesrgan-ncnn-vulkan -i tmp_frames -o out_frames -n realesr-animevideov3 -s 2 -f jpg

  3. Merge the enhanced frames back into a video

    ffmpeg -i out_frames/frame%08d.jpg -i onepiece_demo.mp4 -map 0:v:0 -map 1:a:0 -c:a copy -c:v libx264 -r 23.98 -pix_fmt yuv420p output_w_audio.mp4

If you encounter the problem that macOS cannot verify that this app is free from Malware, you have two options to try:

  1. Choose Apple menu > go to System Preferences > choose Security & Privacy > tap the General tab > click the Open Anyway button.
  2. OR run command: sudo spctl --master-disable to disable Gatekeeper.

GitHub: https://github.com/xinntao/Real-ESRGAN/ Paper: https://arxiv.org/abs/2107.10833

This executable file is portable and includes all the binaries and models required. No CUDA or PyTorch environment is needed.

Note that it may introduce block inconsistency (and also generate slightly different results from the PyTorch implementation), because this executable file first crops the input image into several tiles, and then processes them separately, finally stitches together.

This executable file is based on the wonderful Tencent/ncnn and realsr-ncnn-vulkan by nihui.