Use of an autoencoding model for denoising synthetic images

Jérôme BUISINE 2a9d995fc6 Merge branch 'release/v0.0.4' 1 year ago
generate 138c254d1d Refactoring and scripts updates 1 year ago
modules @ 7e65b752b6 138c254d1d Refactoring and scripts updates 1 year ago
.gitignore b5efdd2af6 Update of ignore files from project 1 year ago
.gitmodules 138c254d1d Refactoring and scripts updates 1 year ago
LICENSE bcc604e079 First model commit 1 year ago
README.md 138c254d1d Refactoring and scripts updates 1 year ago
__init__.py 138c254d1d Refactoring and scripts updates 1 year ago
custom_config.py 138c254d1d Refactoring and scripts updates 1 year ago
dataset 19b3967aa2 Update of model architecture 1 year ago
image_denoising.py 138c254d1d Refactoring and scripts updates 1 year ago
requirements.txt 3fcd2818bf Update of dataset generation 1 year ago

README.md

Denoising with autoencoder

Description

Utilisation d'un autoencoder pour apprendre statistiquement comment il est possible de générer une image de synthèse.

Input :

  • Noisy image
  • Z-buffer
  • Normal card

or other information...

Output :

  • Reference image

Requirements

git clone --recursive https://github.com/prise-3d/Thesis-Denoising-autoencoder.git XXXXX
pip install -r requirements.txt

How to use ?

Autoencoder keras documentation

Generate reconstructed data from specific method of reconstruction (run only once time or clean data folder before):

python generate/generate_reconstructed_data.py -h

Generate custom dataset from one reconstructed method or multiples (implemented later)

python generate/generate_dataset.py -h

Reconstruction parameter (--params)

List of expected parameter by reconstruction method:

  • svd_reconstruction: Singular Values Decomposition
    • Param definition: interval data used for reconstruction (begin, end)
    • Example: "100, 200"
  • ipca_reconstruction: Iterative Principal Component Analysis
    • Param definition: number of components used for compression and batch size
    • Example: "30, 35"
  • fast_ica_reconstruction: Fast Iterative Component Analysis
    • Param definition: number of components used for compression
    • Example: "50"
  • static Use static file to manage (such as z-buffer, normals card...)
    • Param definition: Name of image of scene need to be in {sceneName}/static/xxxx.png
    • Example: "img.png"

Example:

python generate/generate_dataset.py --output data/output_data_filename --metrics "svd_reconstruction, ipca_reconstruction, fast_ica_reconstruction" --renderer "maxwell" --scenes "A, D, G, H" --params "100, 200 :: 50, 10 :: 50" --nb_zones 10 --random 1 --only_noisy 1

Then, run the model:

python image_denoising --data data/my_dataset --output output_model_name

License

The MIT license