Thank you! Finally something relevant. However, I tried through tensorRT and it didn't load the video card much, but more the processor. in Flowframes implementation through VapourSynth even worked faster it seems. Still, the main thing is that there are actual RIFE models here. 👍👍
I was wondering if this is the same tech to get temporally coherent frames for animation. Basically most output for video based AI is jittery . Somehow RunwayML gets this correct. What are they using or doing ? Thanks
@@StableAIHub Just to clarify , I dont think RunwayML has an interpolator as show in the your video sure. what I meant to ask was if there is any underlying tech behind getting temporally consistent results from in their video results. Cheers
Thank you! Finally something relevant. However, I tried through tensorRT and it didn't load the video card much, but more the processor. in Flowframes implementation through VapourSynth even worked faster it seems.
Still, the main thing is that there are actual RIFE models here. 👍👍
Thanks for the vid, finally a good alternative to Flowframes :D
Glad you liked it! Flowframes was good but not updated anymore to include new models.
Awesome, thanks.
u r welcome
Works fine. Good thing is it is updated frequently.
Enjoy!!
I was wondering if this is the same tech to get temporally coherent frames for animation. Basically most output for video based AI is jittery . Somehow RunwayML gets this correct.
What are they using or doing ?
Thanks
I will have to check, never used RunwayML interpolation. Will definitely post if I get to know.
@@StableAIHub Just to clarify , I dont think RunwayML has an interpolator as show in the your video sure. what I meant to ask was if there is any underlying tech behind getting temporally consistent results from in their video results.
Cheers
@@behrampatel4872 Sorry I misunderstood. I unfortunately don't have any knowledge on this. Will try to see and understand how it works.