Skip to content

Latest commit

 

History

History
121 lines (82 loc) · 4.71 KB

anime_video_model.md

File metadata and controls

121 lines (82 loc) · 4.71 KB

Anime Video Models

✅ We add small models that are optimized for anime videos :-)

Models Scale Description
RealESRGANv2-animevideo-xsx2 X2 Anime video model with XS size
RealESRGANv2-animevideo-xsx4 X4 Anime video model with XS size

The following are some demos (best view in the full screen mode).

naruto_X2_compare.MP4
The.Monkey.King_X2_compare.MP4
Fate.stay.night.UBW.X2.Compare.MP4

How to Use

PyTorch Inference

# download model
wget https://github.com/xinntao/Real-ESRGAN/releases/download/v0.2.3.0/RealESRGANv2-animevideo-xsx2.pth -P experiments/pretrained_models
# inference
python inference_realesrgan_video.py -i inputs/video/onepiece_demo.mp4 -n RealESRGANv2-animevideo-xsx2 -s 2 -v -a --half --suffix outx2

ncnn Executable File

Step 1: Use ffmpeg to extract frames from video

ffmpeg -i onepiece_demo.mp4 -qscale:v 1 -qmin 1 -qmax 1 -vsync 0 tmp_frames/frame%08d.png
  • Remember to create the folder tmp_frames ahead

Step 2: Inference with Real-ESRGAN executable file

  1. Download the latest portable Windows / Linux / MacOS executable files for Intel/AMD/Nvidia GPU

  2. Taking the Windows as example, run:

    ./realesrgan-ncnn-vulkan.exe -i tmp_frames -o out_frames -n RealESRGANv2-animevideo-xsx2 -s 2 -f jpg
    • Remember to create the folder out_frames ahead

Step 3: Merge the enhanced frames back into a video

  1. First obtain fps from input videos by

    ffmpeg -i onepiece_demo.mp4
    Usage:
    -i                   input video path

    You will get the output similar to the following screenshot.

  2. Merge frames

    ffmpeg -r 23.98 -i out_frames/frame%08d.jpg -c:v libx264 -r 23.98 -pix_fmt yuv420p output.mp4
    Usage:
    -i                   input video path
    -c:v                 video encoder (usually we use libx264)
    -r                   fps, remember to modify it to meet your needs
    -pix_fmt             pixel format in video

    If you also want to copy audio from the input videos, run:

    ffmpeg -r 23.98 -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
    Usage:
    -i                   input video path, here we use two input streams
    -c:v                 video encoder (usually we use libx264)
    -r                   fps, remember to modify it to meet your needs
    -pix_fmt             pixel format in video

More Demos

  • Input video for One Piece:

    onepiece_demo.mp4
  • Out video for One Piece

    onepiece_demo_outx4.mp4

More comparisons

The.Legend.Of.Nezha_X4_compare.MP4