hjwdzh / adversarialtexture Goto Github PK

View Code? Open in Web Editor NEW

175.0 8.0 18.0 1.42 MB

Adversarial Texture Optimization from RGB-D Scans (CVPR 2020).

Home Page: http://stanford.edu/~jingweih/papers/advtex/

License: MIT License

C++ 34.87% Shell 2.41% Python 47.04% Cuda 13.36% CMake 2.32%

texture adversarial reconstruction 3d-vision cvpr2020

adversarialtexture's Introduction

AdversarialTexture

Adversarial Texture Optimization from RGB-D Scans (CVPR 2020).

Scanning Data Download

Please refer to data directory for details.

Before run following scripts, please modify the data_path in src/config.py as the absolute path of the data folder (e.g. Adversarial/data) where you download all data.

Prepare for Training (Optimization)

Please refer to src/preprocessing directory for details.

Run Training (Optimization)

Consider execute run_all.sh in parallel.

cd src/textureoptim
python gen_script.py
sh run_all.sh

Result Visualization

The result will be stored in data/result/chairID/chairID.png. You can use them to replace the corresponding default texture in data/shape, and use meshlab to open obj files to see the results.

Alternatively, we provide a simple script to render results. You will be able to see the rendering comparison in data/visual.

cd src
python visualize.py

Authors

Jingwei Huang

IMPORTANT: If you use this code please cite the following in any resulting publication:

@inproceedings{huang2020adversarial,
  title={Adversarial Texture Optimization from RGB-D Scans},
  author={Huang, Jingwei and Thies, Justus and Dai, Angela and Kundu, Abhijit and Jiang, Chiyu and Guibas, Leonidas J and Niessner, Matthias and Funkhouser, Thomas},
  booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition},
  pages={1559--1568},
  year={2020}
}

The rendering process is a modification of pyRender.

adversarialtexture's People

Contributors

Stargazers

Watchers

Forkers

peterzs bobiblazeski zebrajack nonlinear1 lotayou fdp0525 peterzhousz kirumang baldrlector wavelet303 wbaileyisla hendriktpl lukashoel hsiangyangchu wenjiar pavan4 babyblue26 gaybro8777

adversarialtexture's Issues

Have you ever seen the `134` error code when calling `Sen.Parse` method?

All data is downloaded from your README file.

And it has already been compiled and passed.

Sens.Parse(c_char_p(filename.encode('utf8')))

Process finished with exit code 134

Cannot download the data

Hello, I tried to download the data with commands in the readme at data directory, and they all response 404.
Is there any thing wrong with the data source?
Expect for your reply!

Very impressive work! While checking the dataset file, I found you utilize the cache to ensure no repeat calculation for each source image. But it will lead to one-to-one matching during training, rather than selecting the target view randomly. Is it right?

OSError when excuting src/preprocessing/run.sh

OSError: ./Rasterizer/libSens.so: undefined symbol: _ZN2cv8imdecodeERKNS_11_InputArrayEi

Is there any suggestion?

thanks

Questions about the code

Dear Jingwei, I have other two questions about the code

In Line128 of dataset.py, why 1-uv?
In Line 146 of model.py, why discrim_loss is set to 0 when it is greater than gen_loss_GAN?

Versions of dependent packages

Hello,could you write the Dependencies in readme?eg. I install TF2 but the code need TF1.I would be very grateful if you could accept my opinion.Thanks.

About Qianyi Zhou's data - the fountain model

It's a great work and thank you for sharing the code!

I have tried to run the code with your chair data, and it did show a significant improvement compared to L1 result.

Then, I edit render_scan.py file to read data from Zhou's fountain model (I choose around 30 RGBD frames as key frames, and use code from "Let there be color" to get the obj and mtl files from all 30 key frames.), I run the code with default parameters (λ=10.0, iter=4001).

But I don't get the fine result as the supplemental shows.

Here's what I get:

Before (L1):

After Iterations:

Is anything missing to run such RGBD datasets?

Thanks again for the interesting work!

About custom dataset

Hi, thank you for sharing source code, it's a fancy work!

I have a question about using custom dasaset. Since I have some RGBD dataset, I composed it to .sens file like BundleFusion (e.g. apt0.sens). But it couldn't work well using render_scan.py.

Any suggestion or is there a detail about how your .sens file composed?

Thank you~

What do you use the coordinate system for `.sens` dataset.

I'm trying to use another method to render your dataset, but the coordinate seems wired.