Dear,
In your paper, it is written like 'Measuring FPS. We compute FPS for all frames of the video and report the median value in our
experiments including the time taken to decode the residuals. Note that the decoding is a one-time operation per step and rendering for a preceding frame can be performed while learning the residuals of a subsequent ones to achieve even higher speeds.'
But in the code, there are no ways to measuring fps like the paper.
I trained coffee_martini use 'log_compressed' and I also downloaded pretrained compressed coffee_martini model.
Is there anything I might have missed? Thank you.
Dear,
In your paper, it is written like 'Measuring FPS. We compute FPS for all frames of the video and report the median value in our
experiments including the time taken to decode the residuals. Note that the decoding is a one-time operation per step and rendering for a preceding frame can be performed while learning the residuals of a subsequent ones to achieve even higher speeds.'
But in the code, there are no ways to measuring fps like the paper.
I trained coffee_martini use 'log_compressed' and I also downloaded pretrained compressed coffee_martini model.
Is there anything I might have missed? Thank you.