search

chungyiweng / humannerf

HumanNeRF turns a monocular video of moving people into a 360 free-viewpoint video.
MIT License
786 stars 86 forks source link

How to put the people in the picture into the volume space?（As shown in the picture blow in the paper） #90

Open luoshuiyue opened 6 months ago

luoshuiyue commented 6 months ago

image

Dipankar1997161 commented 6 months ago

I think that's just a simple diagram. try draw.io

luoshuiyue commented 5 months ago

I think that's just a simple diagram. try draw.io

Thanks for reply! But I still want to know if this is merely a schematic diagram, then where do the three-dimensional point coordinates input for the neural network come from? If they are obtained from estimated human poses, then the key points of the human body seem too sparse. Therefore, I would like to know how the observation space is obtained. Or perhaps my understanding of the paper is mistaken?