Date of recent model update: 2020.11.16, v1.1
You can easily use colaboratory to have a try and generate the youtube demo.
Our model can run 30+FPS for 2X 720p interpolation on a 2080Ti GPU. Currently our method supports 2X/4X interpolation for video, and multi-frame interpolation between a pair of images. Everyone is welcome to use this alpha version and make suggestions!
16X interpolation results from two input images:
We propose a real-time intermediate flow estimation algorithm (RIFE) for video frame interpolation (VFI). Most existing methods first estimate the bi-directional optical flows, and then linearly combine them to approximate intermediate flows, leading to artifacts around motion boundaries. We design an intermediate flow model named IFNet that can directly estimate the intermediate flows from coarse to fine. We then warp the input frames according to the estimated intermediate flows and employ a fusion process to compute final results. Based on our proposed leakage distillation, RIFE can be trained end-to-end and achieve excellent performance. Experiments demonstrate that RIFE is significantly faster than existing flow-based VFI methods and achieves state-of-the-art index on several benchmarks.
$ pip3 install tqdm
$ pip3 install torch
$ pip3 install numpy
$ pip3 install opencv-python
- Download the pretrained models from here. We are optimizing the visual effects and will support animation in the future.
(我们也提供了百度网盘链接:https://pan.baidu.com/s/1YVUsusJFhZ2rWg1Zs5sOkQ 密码:88bu，把压缩包解开后放在 train_log/*.pkl)
- Unzip and move the pretrained parameters to train_log/*.pkl
The models under different setting is coming soon.
Video 2x Interpolation
You can use our demo video or use your own video to run our model.
$ python3 inference_mp4_2x.py --video video.mp4 --fps=60
(generate video_2x.mp4, you can use this script recursively)
$ python3 inference_mp4_4x.py --video video.mp4 --fps=60
(if you want 4x interpolation)
$ python3 inference_mp4_2x.py --video video.mp4 --montage
(if you want to montage the origin video)
The warning info, 'Warning: Your video has *** static frames, it may change the duration of the generated video.' means that your video has changed the frame rate by adding static frames, it is common if you have processed 24FPS video to 30FPS.
Image Interpolation
$ python3 inference_img.py --img img0.png img1.png --times=4
(2^4=16X interpolation results)
$ ffmpeg -r 10 -f image2 -i output/img%d.png -s 448x256 -c:v libx264 -pix_fmt yuv420p output/slomo.mp4 -q:v 0 -q:a 0
(generate a slomo mp4 video based on two input images)
$ ffmpeg -r 10 -f image2 -i output/img%d.png -s 448x256 -vf "split[s0][s1];[s0]palettegen=stats_mode=single[p];[s1][p]paletteuse=new=1" output/slomo.gif
RIFE reported by our paper We will release our training and benchmark validation code soon.
Vimeo90K Download Vimeo90K dataset at ./vimeo_interp_test
$ python3 Vimeo90K_benchmark.py
(You will get 35.695PSNR and 0.9788SSIM)
@article{huang2020rife,
title={RIFE: Real-Time Intermediate Flow Estimation for Video Frame Interpolation},
author={Huang, Zhewei and Zhang, Tianyuan and Heng, Wen and Shi, Boxin and Zhou, Shuchang},
journal={arXiv preprint arXiv:2011.06294},
year={2020}
}
Optical Flow: ARFlow pytorch-liteflownet RAFT
Video Interpolation: DAIN CAIN AdaCoF-pytorch