Skip to content

Commit

Permalink
latest added EHT
Browse files Browse the repository at this point in the history
  • Loading branch information
@pavlosprotopapas
pavlosprotopapas committed Dec 5, 2024
1 parent c4903e5 commit 3eb0f10
Show file tree
Hide file tree
Showing 10 changed files with 65 additions and 33 deletions.
Binary file modified assets/general/.DS_Store
View file
Binary file not shown.
Binary file modified assets/general/img/.DS_Store
View file
Binary file not shown.
Binary file added assets/general/img/projects/MAD_real_deblur.png
View file
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Binary file added ...ts/MAD_real_deblur_M87_REAL_a=0.00_inc=163_Rhigh=000_freq=230_fr100_blur000.jpg
View file
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Binary file added ...ts/MAD_real_deblur_M87_REAL_a=0.00_inc=163_Rhigh=000_freq=230_fr100_blur000.png
View file
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Binary file modified assets/general/img/projects/super-resolution.png
View file
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Binary file added assets/general/img/projects/uvplane.jpeg
View file
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Binary file added assets/general/img/under-construction-gif-17.gif
View file
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
30 changes: 15 additions & 15 deletions people.html
View file
Original file line number Diff line number Diff line change
Expand Up @@ -69,21 +69,21 @@ <h1 class="display-4">Principal Investigator</h1>
src="assets/general/img//home/harvard-university.png"-->
<div class="mb-6">
<blockquote class="h3 font-weight-normal text-lh-lg">
Pavlos is an educator and researcher. As an educator, Pavlos is
teaching <a href="courses.html#CS109A">CS109A</a>,
<a href="courses.html#CS109B">CS109B</a>, introduction to data science and advanced
topics of data science. He also teaches a course in <a href="courses.html#AC215">MLOps</a>.
In the past he has taught <a href="https://www.capstone.iacs.seas.harvard.edu">capstone</a> courses in data science
and computational science, introduction to deep reinforcement
learning, and planning a course in physics informed nuaral
networks. Besides teaching regular courses at Harvard,
his courses are available via the Harvard Extension School,
and <a href="courses.html#HarvardX">HarvardX</a>. He has also participate at data science schools
in <a href="http://lssds.aura-astronomy.org/winter_school/">Chile</a> , Colombia, India and Rwanda. His research
is in the intersection of astronomy, machine learning and
statistics. Most of the research activities are described
in these pages. <br>
You can find his CV <a href="assets/general/JA/CV_Protopapas.pdf">here</a>.
Pavlos is an educator and researcher. As an educator, Pavlos teaches
<a href="courses.html#CS109A">CS109A</a>,
<a href="courses.html#CS109B">CS109B</a>, Introduction to Data Science, and Advanced
Topics in Data Science. He also teaches a course in <a href="courses.html#AC215">MLOps</a>.
In the past, he has taught <a href="https://seas.harvard.edu/applied-computation/courses">Capstone</a> courses in Data Science
and Computational Science, Introduction to Deep Reinforcement
Learning, and is planning a course in Physics-Informed Neural
Networks. Besides teaching regular courses at Harvard,
his courses are also available via the Harvard Extension School
and <a href="courses.html#HarvardX">HarvardX</a>. He has also participated in Data Science schools
in <a href="http://lssds.aura-astronomy.org/winter_school/">Chile</a>, Colombia, India, and Rwanda. His research
lies at the intersection of astronomy, machine learning, and
statistics. Most of his research activities are described
on these pages. <br>
You can find his CV <a href="assets/general/JA/CV_Protopapas.pdf">here</a>.
</blockquote>
</div>

Expand Down
68 changes: 50 additions & 18 deletions projects/nneht.html
View file
Original file line number Diff line number Diff line change
Expand Up @@ -46,7 +46,7 @@ <h1>NN-EHT projects</h1>
</div>
<div class="description" >
<p class="proj-title">Parameterization of the M87* blackhole</p>
<p style="margin-bottom:auto;"> People: <a href="">Tao Tsui</a>, Varshini Reddy, Cecilia Garraffo, Pavlos Protopapas</p>
<p style="margin-bottom:auto;"> People: <a href="">Tao Tsui</a> <em>et al.</em></p></p>
<p style="margin-bottom:auto;">Deep-learning-based computer vision models
have demonstrated superior performance
on a great variety of image related tasks, including image classification
Expand All @@ -59,8 +59,10 @@ <h1>NN-EHT projects</h1>
super-resolution imaging.</p>

<!--a href="https://www.stellardnn.org/" class="link">Webpage</a-->
<a href=" " class="link">Paper</a>
<a href=" " class="link">Code</a>
<a href="../assets/general/img/under-construction-gif-17.gif" class="link">Paper</a>
<!-- <a href="" class="link"> -->
Code
<!-- </a> &nbsp;&nbsp;</p> -->
</div>
</div>
</div>
Expand All @@ -74,7 +76,7 @@ <h1>NN-EHT projects</h1>
</div>
<div class="description" >
<p class="proj-title">Parameterization of the M87* blackhole using Generative Adversarial Networks</p>
<p style="margin-bottom:auto;"> People: <a href="https://www.arya-mohan.com/">Arya Mohan</a>, Pavlos Protopapas</p>
<p style="margin-bottom:auto;"> People: <a href="https://www.arya-mohan.com/">Arya Mohan</a> <em>et al.</em></p></p>
<p style="margin-top:7px; margin-bottom: auto;">Accurate parameterisation of the M87* blackhole is
challenging as the simulations are computationally expensive resulting in
sparse training datasets. In order to increase the size of the training grid,
Expand All @@ -83,29 +85,59 @@ <h1>NN-EHT projects</h1>
variety of synthetic black hole images based on its
spin and electron distribution parameters.</p>
<!--a href="https://www.stellardnn.org/" class="link">Webpage</a-->
<a href="" class="link">Paper</a>
<a href="" class="link">Code</a> &nbsp;&nbsp;</p>
<a href="https://academic.oup.com/mnras/article/527/4/10965/7469481?login=false" class="link">Paper</a>
<!-- <a href="" class="link"> -->
Code
<!-- </a> &nbsp;&nbsp;</p> -->
</div>
</div>
</div>

<br> <br>
<div class="main-container" >
<div class="image-description-container">
<div class="image-container">
<img src="/assets/general/img/projects/NNEHT_GAN.jpeg" class="image" alt="Real and Fake Blackholes">
<img src="/assets/general/img/projects/MAD_real_deblur.png" class="image" alt="Real and Fake Blackholes">
</div>
<div class="description" >
<p class="proj-title">Parameterization of the M87* blackhole using Generative Adversarial Networks</p>
<p style="margin-bottom:auto;"> People: <a href="https://www.arya-mohan.com/">Lily </a>, Pavlos Protopapas</p>
<p style="margin-top:7px; margin-bottom: auto;">Accurate parameterisation of the M87* blackhole is
challenging as the simulations are computationally expensive resulting in
sparse training datasets. In order to increase the size of the training grid,
we propose a data augmentation methodology based
on Conditional Progressive Generative Adversarial Networks to generate a
variety of synthetic black hole images based on its
spin and electron distribution parameters.</p>
<p class="proj-title">Reconstructing M87* Black Hole Images using Multi-Conditional Diffusion Models</p>
<p style="margin-bottom:auto;"> People: <a href="https://www.linkedin.com/in/yuqing-pan/">Yuqing Pan</a> <em>et al.</em></p></p>
<p style="margin-top:7px; margin-bottom: auto;">Diffusion models have become popular neural network architectures
due to their success in various computer vision applications. This paper bridges the gap between advanced
deep learning techniques and astronomical imagery. We present a multi-conditional diffusion model,
InstructPix2Pix-M87, trained on general relativistic magnetohydrodynamic (GRMHD) simulated images of the
M87* black hole to enhance the quality of observations from the Event Horizon Telescope (EHT).
Our approach demonstrates improved de-blurring and parameter inference capabilities for M87*,
laying the groundwork for further studies in astrophysics..</p>
<!--a href="https://www.stellardnn.org/" class="link">Webpage</a-->
<a href="../assets/general/img/under-construction-gif-17.gif" class="link">Paper</a>
<!-- <a href="" class="link"> -->
Code
<!-- </a> &nbsp;&nbsp;</p> -->
</div>
</div>
</div>

<br> <br>
<div class="main-container" >
<div class="image-description-container">
<div class="image-container">
<img src="/assets/general/img/projects/uvplane.jpeg" class="image" alt="Real and Fake Blackholes">
</div>
<div class="description" >
<p class="proj-title">Parametrization of M87* in the u-v visibility spectrum</p>
<p style="margin-bottom:auto;"> People: Fanc O <em>et al.</em></p></p>
<p style="margin-top:7px; margin-bottom: auto;"><p>The Event Horizon Telescope (EHT) has revolutionized black hole
physics by enabling direct imaging of supermassive black holes like M87* and Sgr A*. This research explores deep
learning methods to analyze M87* data in the frequency domain, bypassing image reconstruction.
We evaluated neural network architectures for parameter inference, training on SANE and MAD simulations.
Static and dynamic models performed well for SANE, but MAD required dynamic or polarimetric data for accurate
results. Our analysis of real M87* data demonstrates the promise of this novel approach.</p></p>
<!--a href="https://www.stellardnn.org/" class="link">Webpage</a-->
<a href="" class="link">Paper</a>
<a href="" class="link">Code</a> &nbsp;&nbsp;</p>
<a href="../assets/general/img/under-construction-gif-17.gif" class="link">Paper</a>
<!-- <a href="" class="link"> -->
Code
<!-- </a> &nbsp;&nbsp;</p> -->
</div>
</div>
</div>
Expand Down

0 comments on commit 3eb0f10

Please sign in to comment.