Name		Name	Last commit message	Last commit date
Latest commit History 2 Commits
ContextualBandit.py		ContextualBandit.py
ContextualBanditDiscreteActionSpace.py		ContextualBanditDiscreteActionSpace.py
FourierOptics.py		FourierOptics.py
Global_piston_restoration_of_segmented_mirrors_with_convolutional_recurrent_neural_network_approach.pdf		Global_piston_restoration_of_segmented_mirrors_with_convolutional_recurrent_neural_network_approach.pdf
Mirror.py		Mirror.py
MountainCar.py		MountainCar.py
MountainPiston.py		MountainPiston.py
PistonCNN.py		PistonCNN.py
PistonTools.py		PistonTools.py
README.md		README.md
SVD_Actor_Critic.py		SVD_Actor_Critic.py
TFCrossvalidation.py		TFCrossvalidation.py
TFInference.py		TFInference.py
TFReadRecords.py		TFReadRecords.py
TFRegressionFromUncompleteData.py		TFRegressionFromUncompleteData.py
TFThreads.py		TFThreads.py
TFTrain.py		TFTrain.py
TFTrainCNN.py		TFTrainCNN.py
TFTrainCNNFromMemory.py		TFTrainCNNFromMemory.py
TFTrainLinearCNNFromMemory.py		TFTrainLinearCNNFromMemory.py
TFTrainLinearRegressionFromMemory.py		TFTrainLinearRegressionFromMemory.py
TestSVD.py		TestSVD.py
i1.png		i1.png
my_test.py		my_test.py
ol-43-17-4264.pdf		ol-43-17-4264.pdf
pistonREINFORCEMENT.py		pistonREINFORCEMENT.py
pistonRNN.py		pistonRNN.py
simulacion_piston.py		simulacion_piston.py
test2.py		test2.py
testActorCritic.py		testActorCritic.py
test_ps.py		test_ps.py
valid_first_test.py		valid_first_test.py
vray_osl_simplex_noise_10.png		vray_osl_simplex_noise_10.png
wf1.png		wf1.png
wf2.png		wf2.png

Repository files navigation

pistonCNN

About

Measurement of mirror piston misalignments for large segmented telescopes. Diffraction images are generated to train neural networks. The goal is to quantify non-smooth wavefront phase shifts. Various deep learning approaches are used: CNN, RNN and reinforcement learning.

Readme