Performance on multiview setting

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

1.75k stars 342 forks source link

Performance on multiview setting #108

Open sta105 opened 3 years ago

sta105 commented 3 years ago

Hi shunsuke,

I trained PIFU under multiview setting(4 random views) on twindom dataset, and tested its performance on renderpeople rigged data(I used your data generation code with the same rendering setting to render twindom and renderpeople models). However, multiview-pifu seems only work on standing poses. like:

test_eval_epoch0_000001

and:

if the pose changes, the results will be very bad:

Does Multiview PIFU overfit on standing poses?

EthanZhangCN commented 2 years ago

hello, can you share how to train by multi-view?

caiyongqi commented 2 years ago

Hi shunsuke,

I trained PIFU under multiview setting(4 random views) on twindom dataset, and tested its performance on renderpeople rigged data(I used your data generation code with the same rendering setting to render twindom and renderpeople models). However, multiview-pifu seems only work on standing poses. like:

and:

if the pose changes, the results will be very bad:

Does Multiview PIFU overfit on standing poses?

hello, can you share how to train by multi-view?