Skip to content

Latest commit

 

History

History

tvsynth

Folders and files

NameName
Last commit message
Last commit date

parent directory

..
README.md
README.md
 
 
config.py
config.py
 
 
config_robustness.py
config_robustness.py
 
 
data_management.py
data_management.py
 
 
find_adversarial.py
find_adversarial.py
 
 
networks.py
networks.py
 
 
operators.py
operators.py
 
 
reconstruction_methods.py
reconstruction_methods.py
 
 
script_robustness_example_adv.py
script_robustness_example_adv.py
 
 
script_robustness_example_bernoulli.py
script_robustness_example_bernoulli.py
 
 
script_robustness_example_gauss.py
script_robustness_example_gauss.py
 
 
script_robustness_example_uniform.py
script_robustness_example_uniform.py
 
 
script_robustness_table_adv.py
script_robustness_table_adv.py
 
 
script_robustness_table_bernoulli.py
script_robustness_table_bernoulli.py
 
 
script_robustness_table_gauss.py
script_robustness_table_gauss.py
 
 
script_robustness_table_uniform.py
script_robustness_table_uniform.py
 
 
script_train_tiramisu.py
script_train_tiramisu.py
 
 
script_train_tiramisu_ee.py
script_train_tiramisu_ee.py
 
 
script_train_tiramisu_ee_jitter.py
script_train_tiramisu_ee_jitter.py
 
 
script_train_tiramisu_jitter.py
script_train_tiramisu_jitter.py
 
 
script_train_unet.py
script_train_unet.py
 
 
script_train_unet_ee.py
script_train_unet_ee.py
 
 
script_train_unet_ee_jitter.py
script_train_unet_ee_jitter.py
 
 
script_train_unet_it_tikh.py
script_train_unet_it_tikh.py
 
 
script_train_unet_it_tikh_jitter.py
script_train_unet_it_tikh_jitter.py
 
 
script_train_unet_jitter.py
script_train_unet_jitter.py
 
 

README.md

tvsynth

Signal recovery of piecewise constant 1D signals (following a total variation synthesis model) from random Gaussian measurements.

How to run this experiment

  1. Check (and modify if necessary) the configuration file config.py. It specifies the directory paths for the data and results, as well as all relevant parameters of the experimental setup. By default, the data is stored in the subdirectory raw_data and results and model weights are stored in the subdirectory results.
  2. Generate and prepare the data by running data_management.py.
  3. Train networks using the scripts named script_train_*.py.
  4. Check (and modify if necessary) the configuration file config_robustness.py. It specifies the relevant parameters for the robustness analysis. Adapt the list of networks to compare according to the ones that you have actually trained.
  5. Analyze different aspects of the robustness comparisons by running the scripts named script_robustness_*.py.