Statistical analysis of sample values to approximate the final pixel value

Jérôme BUISINE 4055b069ea Merge branch 'release/v0.1.1' 8 months ago
generate 95cba96b9b Use of modules dependency 10 months ago
modules @ d5de038bdc 95cba96b9b Use of modules dependency 10 months ago
notebooks 87b8d18661 Write Keras result script 1 year ago
others 95cba96b9b Use of modules dependency 10 months ago
reconstruct 95cba96b9b Use of modules dependency 10 months ago
.gitignore ab86ef6206 Documentation and ignore files updates 10 months ago
.gitmodules 95cba96b9b Use of modules dependency 10 months ago
LICENSE d0ebc63822 Update of documentation 11 months ago
README.md ab86ef6206 Documentation and ignore files updates 10 months ago
__init__.py beb4716c41 Add of feature choices 10 months ago
compare_images.py 95cba96b9b Use of modules dependency 10 months ago
custom_config.py adf8908549 Update of keras config 8 months ago
features.py 95cba96b9b Use of modules dependency 10 months ago
generate_data.sh cccee3b412 Update of Keras scripts 1 year ago
run.sh 8590bd14ac Update of run script 9 months ago
run_keras.sh a6aa09a009 Keras run script update 9 months ago
train_model.py 018d8e801f Update of train model script 10 months ago
train_model_keras.py 95cba96b9b Use of modules dependency 10 months ago

README.md

Sample Analysis

Description

The aim of this project is to predict the mean pixel value from monte carlo process rendering in synthesis images using only few samples information in input for model.

Data

Data are all scenes samples information obtained during the rendering process.

For each pixel we have a list of all grey value estimated (samples).

Models

List of models tested :

  • Ridge Regression
  • SGD
  • SVR (with rbf kernel)

How to use

First you need to contact jerome.buisine@univ-littoral.fr in order to get datatset version. The dataset is not available with this source code.

python make_dataset.py --n 10 --each_row 8 --each_column 8
python reconstruct.py --scene Scene1 --model_path saved_models/Model1.joblib --n 10 --image_name output.png

License

The MIT license