RIFE: Real-Time Intermediate Flow Estimation for Video Frame Interpolation

Creates sequence of interpolated frames between given input images

Run

python interpolate.py --input samples/ --output frames/ --buffer 25 --multi 25

interpolating 4 images
image samples/image000.jpg ssim 0.99 buffer 25 frames
image samples/image001.jpg ssim 0.54 create 69 frames
image samples/image002.jpg ssim 0.45 create 69 frames
image samples/image003.jpg ssim 0.55 create 69 frames
image samples/image003.jpg ssim 0.99 buffer 25 frames
frames 259 time 4.24

• Reads input images from samples/ and writes output images to frames/
• Number of generated frames will be 70x input frames
• Start and end will be buffered/padded with 25 frames

ffmpeg -hide_banner -loglevel warning -hwaccel auto -y -framerate 30 -i "frames/%6d.jpg" -r 30 -vcodec libx264 -preset medium -crf 23 -vf minterpolate=mi_mode=blend,fifo -movflags +faststart samples/video.mp4

• Creates a video file from interpolated frames

Options

./interpolate.py --help

--model MODEL    path to model
--input INPUT    input directory containing images
--output OUTPUT  output directory for interpolated images
--scale SCALE    scale factor for interpolated images
--multi MULTI    number of frames to interpolate between two input images
--buffer BUFFER  number of frames to buffer on scene change
--change CHANGE  scene change threshold (lower is more sensitive
--fp16           use float16 precision instead of float32

Example

Both examples are created using SD.Next

Using AnimateDiff extension

rife.mp4

Video: 2.5sec at 25fps using 16 input images

Using Seed Travel extension

video.mp4

Video: 9sec at 30fps using 10 input images

Credits

• https://github.com/megvii-research/ECCV2022-RIFE
• https://github.com/hzwer/Practical-RIFE