Open bigface008 opened 4 years ago
Not for now. You may try SVP tool for real-time application.
多谢!(另外,顺便问下,您知不知道有开源的、利用机器学习进行实时补帧的工具?或者说除了vapoursynth 之外的不用机器学习的可以用来做实时补帧的开源库(或者工具)?(毕竟 SVP 是专有的) Thank you so much! By the way, could you please tell me if there is any open-source tool used for real-time interpolation that use machine learning to work? Excluding vapoursynth, Is there any open-source tool don't use machine learing? (After all, SVP is proprietory.
多谢!(另外,顺便问下,您知不知道有开源的、利用机器学习进行实时补帧的工具?或者说除了vapoursynth 之外的不用机器学习的可以用来做实时补帧的开源库(或者工具)?(毕竟 SVP 是专有的) Thank you so much! By the way, could you please tell me if there is any open-source tool used for real-time interpolation that use machine learning to work? Excluding vapoursynth, Is there any open-source tool don't use machine learing? (After all, SVP is proprietory.
hi,I have same question.Do you find the code about it. Thanks!
多谢!(另外,顺便问下,您知不知道有开源的、利用机器学习进行实时补帧的工具?或者说除了vapoursynth 之外的不用机器学习的可以用来做实时补帧的开源库(或者工具)?(毕竟 SVP 是专有的) Thank you so much! By the way, could you please tell me if there is any open-source tool used for real-time interpolation that use machine learning to work? Excluding vapoursynth, Is there any open-source tool don't use machine learing? (After all, SVP is proprietory.
hi,I have same question.Do you find the code about it. Thanks!
(我就不用散装英语了)目前我的解决方案是用OpenCV的拓展包opencv_contrib里面的optflow和cudaoptflow来计算光流场(cv::optflow::createOptFlow_SparseToDense可以算是速度比较快的,可以做到实时的效果),然后手动地根据光流场计算出新的帧(照理说opencv里应该也有直接根据光流场算新帧的库)。这是传统算法,没有用机器学习。
示例代码可以看看https://github.com/Meltz014/ocv。
至于,机器学习类的实时补帧,我还没找到......
多谢!(另外,顺便问下,您知不知道有开源的、利用机器学习进行实时补帧的工具?或者说除了vapoursynth 之外的不用机器学习的可以用来做实时补帧的开源库(或者工具)?(毕竟 SVP 是专有的) Thank you so much! By the way, could you please tell me if there is any open-source tool used for real-time interpolation that use machine learning to work? Excluding vapoursynth, Is there any open-source tool don't use machine learing? (After all, SVP is proprietory.
hi,I have same question.Do you find the code about it. Thanks!
(我就不用散装英语了)目前我的解决方案是用OpenCV的拓展包opencv_contrib里面的optflow和cudaoptflow来计算光流场(cv::optflow::createOptFlow_SparseToDense可以算是速度比较快的,可以做到实时的效果),然后手动地根据光流场计算出新的帧(照理说opencv里应该也有直接根据光流场算新帧的库)。这是传统算法,没有用机器学习。
示例代码可以看看https://github.com/Meltz014/ocv。
至于,机器学习类的实时补帧,我还没找到......
给力啊,大佬
多谢!(另外,顺便问下,您知不知道有开源的、利用机器学习进行实时补帧的工具?或者说除了vapoursynth 之外的不用机器学习的可以用来做实时补帧的开源库(或者工具)?(毕竟 SVP 是专有的) Thank you so much! By the way, could you please tell me if there is any open-source tool used for real-time interpolation that use machine learning to work? Excluding vapoursynth, Is there any open-source tool don't use machine learing? (After all, SVP is proprietory.
hi,I have same question.Do you find the code about it. Thanks!
(我就不用散装英语了)目前我的解决方案是用OpenCV的拓展包opencv_contrib里面的optflow和cudaoptflow来计算光流场(cv::optflow::createOptFlow_SparseToDense可以算是速度比较快的,可以做到实时的效果),然后手动地根据光流场计算出新的帧(照理说opencv里应该也有直接根据光流场算新帧的库)。这是传统算法,没有用机器学习。 示例代码可以看看https://github.com/Meltz014/ocv。 至于,机器学习类的实时补帧,我还没找到......
给力啊,大佬
:joy:
多谢!(另外,顺便问下,您知不知道有开源的、利用机器学习进行实时补帧的工具?或者说除了vapoursynth 之外的不用机器学习的可以用来做实时补帧的开源库(或者工具)?(毕竟 SVP 是专有的)
Thank you so much! By the way, could you please tell me if there is any open-source tool used for real-time interpolation that use machine learning to work? Excluding vapoursynth, Is there any open-source tool don't use machine learing? (After all, SVP is proprietory.
hi,I have same question.Do you find the code about it. Thanks!
(我就不用散装英语了)目前我的解决方案是用OpenCV的拓展包opencv_contrib里面的optflow和cudaoptflow来计算光流场(cv::optflow::createOptFlow_SparseToDense可以算是速度比较快的,可以做到实时的效果),然后手动地根据光流场计算出新的帧(照理说opencv里应该也有直接根据光流场算新帧的库)。这是传统算法,没有用机器学习。
示例代码可以看看https://github.com/Meltz014/ocv。
至于,机器学习类的实时补帧,我还没找到......
机器学习实时补帧不可能的,至少现在不可能,你看dain补一帧少说也要好几秒,实时一秒至少要补24帧,除非代码优化到极致(不太现实)或者你去搞个超算计算机。
多谢!(另外,顺便问下,您知不知道有开源的、利用机器学习进行实时补帧的工具?或者说除了vapoursynth 之外的不用机器学习的可以用来做实时补帧的开源库(或者工具)?(毕竟 SVP 是专有的) Thank you so much! By the way, could you please tell me if there is any open-source tool used for real-time interpolation that use machine learning to work? Excluding vapoursynth, Is there any open-source tool don't use machine learing? (After all, SVP is proprietory.
hi,I have same question.Do you find the code about it. Thanks!
(我就不用散装英语了)目前我的解决方案是用OpenCV的拓展包opencv_contrib里面的optflow和cudaoptflow来计算光流场(cv::optflow::createOptFlow_SparseToDense可以算是速度比较快的,可以做到实时的效果),然后手动地根据光流场计算出新的帧(照理说opencv里应该也有直接根据光流场算新帧的库)。这是传统算法,没有用机器学习。
示例代码可以看看https://github.com/Meltz014/ocv。
至于,机器学习类的实时补帧,我还没找到......
你好,你的这个代码能公开一下吗?你之前分享那个页面不在了
抱歉打扰了(如果您不介意的话请忽略下面糟糕的中式英语),请问DAIN是否适合实时补帧的场景?如果不适合的话,请您知不知道有别的适合实时补帧、资源占用较低的开源工具呢?我目前急需这类工具,之前看到过您在CVPR上有发表过一篇描述异构系统上进行实时补帧的技术的论文,所以就顺着找了过来,希望能咨询一下您相关领域的东西。我之前有用bigface00@sjtu.edu.cn这一地址向您的gmail邮箱发送过咨询的邮件,不知道您是否有收到?如果您不介意的话,可以告诉我您的QQ或者微信号吗?(或者直接用邮件回复?) 如有打扰,敬请原谅。 Sorry to bother you, but I really need to know whether this tool is suitable for real-time video interpolation. I'm working on my graduation project which need a real-time video interpolation tool. I find that you have published a paper named "High-quality and real-time frame interpolation on heterogeneous computing system" on CVPR, which seems to be a suitable tool but I didn't find its open-source code. Do you know any tools suitable for real-time video interpolation if this project is not what I'm trying to find? 😅