search

shunsukesaito / PIFu

This repository contains the code for the paper "PIFu: Pixel-Aligned Implicit Function for High-Resolution Clothed Human Digitization"
https://shunsukesaito.github.io/PIFu/
Other
1.76k stars 341 forks source link

Training data #47

Closed MC-E closed 4 years ago

MC-E commented 4 years ago

Sorry to bother you, I want to know how many high-resolution photogrammetry scans from RenderPeople and BUFF needed to train this novel model. I try to train this model with only one photogrammetry scan from RenderPeople which is free to download and get a dissatisfied result.

shunsukesaito commented 4 years ago

BUFF is for evaluation only. We used about 460 scans from RenderPeople for training.

MC-E commented 4 years ago

OK, thanks for your reply!

Brawie commented 4 years ago

Hi shunsuke, may I ask which type of meshes from RenderPeople you used, the 'POSED' ones or the 'RIGGED' ones? Are there any other selecting rules or experiences?

shunsukesaito commented 4 years ago

This is the full set of training data. They are POSED model not RIGGED. https://github.com/facebookresearch/pifuhd/blob/master/data/RenderPeople_all.csv