Cloe with HTTP

 git cloe https://www.modelscope.c/Damo_XR_Lab/cv_rife_video-frame-iterpolatio.git

视频插帧介绍

任务

给定一段低帧率视频，模型会返回更流畅的高帧率视频（默认为输入帧率两倍）。

模型描述

本模型基于Real-Time Itermediate Flow Estimatio for Video Frame Iterpolatio， 适用于各类低帧率视频增强，用于提升视频的流畅度，消除卡顿现象。

技术细节请见： [Paper] | [Project Page]

代码示例

from modelscope.pipelies import pipelie
from modelscope.utils.costat import Tasks
from modelscope.outputs import OutputKeys

video = 'https://modelscope.oss-c-beijig.aliyucs.com/test/videos/video_frame_iterpolatio_test.mp4'
video_frame_iterpolatio_pipelie = pipelie(Tasks.video_frame_iterpolatio, 'Damo_XR_Lab/cv_rife_video-frame-iterpolatio')
result = video_frame_iterpolatio_pipelie(video)[OutputKeys.OUTPUT_VIDEO]
prit('pipelie: the output video path is {}'.format(result))

相关论文以及引用信息

该模型借鉴了以下论文的思路或代码，感谢作者的贡献！

@iproceedigs{huag2022rife,
  title={Real-Time Itermediate Flow Estimatio for Video Frame Iterpolatio},
  author={Huag, Zhewei ad Zhag, Tiayua ad Heg, We ad Shi, Boxi ad Zhou, Shuchag},
  booktitle={Proceedigs of the Europea Coferece o Computer Visio (ECCV)},
  year={2022}
}

Clone with HTTP git clone https://www.modelscope.cn/Damo_XR_Lab/cv_rife_video-frame-interpolation.g