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

Questions about texture and multiview #17

Closed nvdomidi closed 4 years ago

nvdomidi commented 4 years ago

Hello shunsuke, I have 3 questions:

  1. After running the code, a colored image like the input is saved. Is this a render of the reconstructed 3D shape or is it just a copy of the input image?
  2. The reconstructed objects I get are without material. Is there an option to save the color?
  3. I cant seem to get the multiview part of the code to work. I set --num_views to the number of pictures I have but it gives errors with tensor sizes.

Great work and thank you in advance!

shunsukesaito commented 4 years ago

After running the code, a colored image like the input is saved. Is this a render of the reconstructed 3D shape or is it just a copy of the input image?

The image is just the visualization of input fed into the network.

The reconstructed objects I get are without material. Is there an option to save the color?

By default, the code generates colored obj files. As the color is assigned per vertex not as a form of texture map, I recommend checking the results with meshlab to make sure color is there.

I cant seem to get the multiview part of the code to work. I set --num_views to the number of pictures I have but it gives errors with tensor sizes.

Please refer to #3. The released model is not trained for multi-view setup and at this moment, we don't plan to release the model for multi-view or example codes (but the code base can be naturally extended to multi-view setup.)

nvdomidi commented 4 years ago

thank you!