Syntesis images noise detection using CNN approach
Jérôme BUISINE 0c91d181de update log information | 3 years ago | |
---|---|---|
analysis | 4 years ago | |
display | 5 years ago | |
generate | 3 years ago | |
models | 5 years ago | |
prediction | 4 years ago | |
run | 4 years ago | |
simulation | 5 years ago | |
.gitignore | 5 years ago | |
LICENSE | 5 years ago | |
README.md | 3 years ago | |
__init__.py | 5 years ago | |
cnn_models.py | 4 years ago | |
config.py | 3 years ago | |
prediction_model.py | 5 years ago | |
requirements.txt | 5 years ago | |
train_lstm_weighted.py | 3 years ago | |
train_model.py | 4 years ago |
git clone --recursive https://github.com/prise-3d/Thesis-NoiseDetection-CNN.git
pip install -r requirements.txt
modules/config/global_config.py
run/runAll_*.sh
script.model_comparisons.csv
file used for store models performance.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
List of expected parameter by reconstruction method:
Example:
python generate/generate_dataset_sequence_file.py --output data/output_data_filename --folder <generated_data_folder> --features "svd_reconstruction, ipca_reconstruction, fast_ica_reconstruction" --params "100, 200 :: 50, 10 :: 50" --sequence 5 --size "100, 100" --selected_zones <zones_files.csv>
Then, train model using your custom dataset:
python train_lstm_model.py --train data/custom_dataset.train --test data/custom_dataset.test --chanels "1,3,3" --epochs 30 --batch_size 64 --seq_norm 1 --output output_model_name
Now we have a model trained, we can use it with an image as input:
python prediction/predict_noisy_image.py --image path/to/image.png --model saved_models/xxxxxx.json --features 'svd_reconstruction' --params '100, 200'
The model will return only 0 or 1:
All scripts named prediction/predict_seuil_expe*.py are used to simulate model prediction during rendering process.
Once you have simulation done. Checkout your threshold_map/%MODEL_NAME%/simulation_curves_zones_*/ folder and use it with help of display_simulation_curves.py script.