update doc to pgann

AmirMardan · Nov 8, 2024 · aa55869 · aa55869
1 parent 9b675ae
commit aa55869
Show file tree

Hide file tree

Showing 3 changed files with 39 additions and 34 deletions.
diff --git a/docs/conf.py b/docs/conf.py
@@ -7,7 +7,7 @@
 # https://www.sphinx-doc.org/en/master/usage/configuration.html#project-information
 
 # don't forget to change html_context['github_repo'] in bottom of this file
-project = 'PINN-FWI'
+project = 'PGNN-FWI'
 copyright = '2023, Amir Mardan'
 author = 'Amir Mardan'
 release = '0.1.0'
@@ -133,6 +133,6 @@
 
 html_context['display_github'] = True
 html_context['github_user'] = 'amirmardan'
-html_context['github_repo'] = 'pinn_fwi'
+html_context['github_repo'] = 'pgnn_fwi'
 html_context['github_version'] = 'main/docs/'
 
diff --git a/docs/index.rst b/docs/index.rst
@@ -1,12 +1,12 @@
-.. PINN-FWI documentation master file, created by
+.. PGNN-FWI documentation master file, created by
    sphinx-quickstart on Thu Mar 28 00:06:12 2024.
    You can adapt this file completely to your liking, but it should at least
    contain the root `toctree` directive.
 
-PIhysics-Informed Neural Networks for Full-Waveform Inversion (PINN-FWI)
+Physics-Guided Neural Networks for Full-Waveform Inversion (PGNN-FWI)
 ========================================================================
 
-PINN-FWI
+PGNN-FWI
 =========
 
 .. figure:: /readme_files/25231.png
@@ -15,10 +15,10 @@ PINN-FWI
    :target: https://github.com/AmirMardan/pinn_fwi
 
 ..caption::
-This is preliminary documentation for Physics-Informed Neural Networks for Full-Waveform Inversion (PINN-FWI)
+This is preliminary documentation for Physics-Guided Neural Networks for Full-Waveform Inversion (PGNN-FWI)
 
-In this repository, I implemented the physics-informed neural network
-(PINN) for full-waveform inversion. This PINN can be implemented with or
+In this repository, I implemented the physics-guided neural network
+(PGNN) for full-waveform inversion. This PGNN can be implemented with or
 without attention block. The architecture of their study is shown in the
 following figure.
 
@@ -46,10 +46,10 @@ specified versions to be sure everything works.
    pip install -r requirements.txt
 
 **In this repo, there are four scripts for running FWI:** 
- | 1. `pinn_fwi.py <https://github.com/AmirMardan/pinn_fwi/blob/main/pinn_fwi.py>`__ for performing PINN- or PIANN-FWI. 
+ | 1. `pinn_fwi.py <https://github.com/AmirMardan/pinn_fwi/blob/main/pinn_fwi.py>`__ for performing PGNN- or PGANN-FWI. 
  | 2. `original_fwi.py <https://github.com/AmirMardan/pinn_fwi/blob/main/original_fwi.py>`__ for running the conventional FWI (Not available). 
- | 3. `pinn_for_init.py <https://github.com/AmirMardan/pinn_fwi/blob/main/pinn_for_init.py>`__ for performing PINN- or PIANN-FWI to create an initial model and use that to perform the conventional FWI (Not available). 
- | 4. `pinn_fwi.ipynb <https://github.com/AmirMardan/pinn_fwi/blob/main/pinn_fwi.ipynb>`__ for performing PINN- or PIANN-FWI, but this notebook might not be updated.
+ | 3. `pinn_for_init.py <https://github.com/AmirMardan/pinn_fwi/blob/main/pinn_for_init.py>`__ for performing PGNN- or PGANN-FWI to create an initial model and use that to perform the conventional FWI (Not available). 
+ | 4. `pinn_fwi.ipynb <https://github.com/AmirMardan/pinn_fwi/blob/main/pinn_fwi.ipynb>`__ for performing PGNN- or PGANN-FWI, but this notebook might not be updated.
 
 The result of running this code for 22 shots with 2500 epochs on the
 Marmousi model is shown in the following figures.
@@ -63,7 +63,7 @@ is using the PIANN-FWI for creating only initial model.
 
 .. toctree::
    :maxdepth: 1
-   :caption: PINN-FWI
+   :caption: PGNN-FWI
 
    source/readme
 
@@ -78,8 +78,8 @@ References:
 
 ::
 
-    @inproceedings{mardan2024piann_eage,
-        title = {Physics-informed attention-based neural networks for full-waveform inversion},
+    @inproceedings{mardan2024pgann_eage,
+        title = {Physics-guided attention-based neural networks for full-waveform inversion},
         author = {Mardan, Amir and Fabien-Ouellet, Gabriel},
         year = {2024},
         booktitle = {85$^{th}$ {EAGE} Annual Conference \& Exhibition},

diff --git a/docs/source/readme.rst b/docs/source/readme.rst
@@ -1,17 +1,9 @@
-PIANN-FWI for estimating the Marmousi velocity model
-====================================================
+PGANN-FWI for estimating the Marmousi velocity model
+===================================================
 
-.. figure:: /readme_files/25231.png
-   :width: 20
-   :alt: git
-   :target: https://github.com/AmirMardan/pinn_fwi
-
-
-This is preliminary documentation for Physics-Informed Neural Networks for Full-Waveform Inversion (PINN-FWI)
-
-In this repository, I implemented the physics-informed neural network
-(PINN) for full-waveform inversion. This PINN can be implemented with or
-without attention block. The architecture of their study is shown in the
+In this repository, I implemented the physics-guided neural network
+(PGNN) for full-waveform inversion. This PGNN can be implemented with or
+without an attention block. The architecture of their study is shown in the
 following figure.
 
 .. figure:: /readme_files/architecture.png
@@ -37,18 +29,31 @@ specified versions to be sure everything works.
 
    pip install -r requirements.txt
 
-**In this repo, there are four scripts for running FWI:** 
- | 1. `pinn_fwi.py <https://github.com/AmirMardan/pinn_fwi/blob/main/pinn_fwi.py>`__ for performing PINN- or PIANN-FWI. 
+In this repo, there are four scripts for running FWI: 
+ | 1. `pinn_fwi.py <https://github.com/AmirMardan/pinn_fwi/blob/main/pinn_fwi.py>`__ for performing PGNN- or PGANN-FWI. 
  | 2. `original_fwi.py <https://github.com/AmirMardan/pinn_fwi/blob/main/original_fwi.py>`__ for running the conventional FWI (Not available). 
- | 3. `pinn_for_init.py <https://github.com/AmirMardan/pinn_fwi/blob/main/pinn_for_init.py>`__ for performing PINN- or PIANN-FWI to create an initial model and use that to perform the conventional FWI (Not available). 
- | 4. `pinn_fwi.ipynb <https://github.com/AmirMardan/pinn_fwi/blob/main/pinn_fwi.ipynb>`__ for performing PINN- or PIANN-FWI, but this notebook might not be updated.
+ | 3. `pinn_for_init.py <https://github.com/AmirMardan/pinn_fwi/blob/main/pinn_for_init.py>`__ for performing PINN- or PGANN-FWI to create an initial model and use that to perform the conventional FWI (Not available). 
+ | 4. `pinn_fwi.ipynb <https://github.com/AmirMardan/pinn_fwi/blob/main/pinn_fwi.ipynb>`__ for performing PINN- or PGANN-FWI, but this notebook might not be updated.
 
-The result of running this code for 22 shots with 2500 epochs on the
-Marmousi model is shown in the following figures.
+The result of running this code for 22 shots with 2500 epochs on the Marmousi model is shown in the following figures.
 
 |res| For a faster convergence (300 epochs), I considered geophones
 around the model and the results are |with_init| where the hybrid method
-is using the PIANN-FWI for creating only initial model.
+is using the PGANN-FWI for creating only the initial model.
+
+Reference:
+
+::
+
+   @inproceedings{mardan2024pgann_eage,
+       title = {Physics-guided attention-based neural networks for full-waveform inversion},
+       author = {Mardan, Amir and Fabien-Ouellet, Gabriel},
+       year = {2024},
+       booktitle = {85$^{th}$ {EAGE} Annual Conference \& Exhibition},
+       publisher = {European Association of Geoscientists \& Engineers},
+       pages = {1-5},
+       doi = {}
+   }
 
 .. |res| image:: /readme_files/marmousi_clean.png
 .. |with_init| image:: /readme_files/image2024_marmousi_clean.png