Ali Siahkoohi
Simons Postdoctoral Fellow
Department of Computational Applied Mathematics and Operations Research
Rice University
Houston, TX, USA
[alisk@rice.edu] [CV] [GitHub] [Google Scholar]
On the academic job market for Fall 2025
Research interests
My research, under the mentorship of Maarten V. de Hoop and Richard G. Baraniuk, focuses on understanding and addressing AI robustness concerns, particularly by designing scalable methods for quantifying uncertainty in AI models through the use of generative models, variational inference, and hypernetworks.
Education
Georgia Institute of Technology, Atlanta, GA, USA
Ph.D., 2022, Computational Science and Engineering
Minor in Applied Mathematics
University of Tehran, Tehran, Iran
M.Sc., 2016, Geophysics
Sharif University of Technology, Tehran, Iran
B.Sc., 2013, Electrical Engineering
Publications
Preprints
Improving fairness and mitigating MADness in generative models
P. M. Mayer, L. Luzi, A. Siahkoohi, D. H. Johnson, and R. G. Baraniuk
2024.
[pdf] [bib]
Martian time-series unraveled: A multi-scale nested approach with factorial variational autoencoders
A. Siahkoohi, R. Morel, R. Balestriero, E. Allys, G. Sainton, T. Kawamura, and M. V. de Hoop
2024.
[pdf] [code] [slides] [bib]
ASPIRE: Iterative amortized posterior inference for Bayesian inverse problems
R. Orozco, A. Siahkoohi, M. Louboutin, and F. J. Herrmann
2024.
[pdf] [code] [bib]
Journal papers
InvertibleNetworks.jl: A Julia package for scalable normalizing flows
R. Orozco, P. Witte, M. Louboutin, A. Siahkoohi, G. Rizzuti, B. Peters, and F. J. Herrmann
Journal of Open Source Software, 9(99):6554, 2024.
[pdf] [code] [link] [bib]
Boomerang: Local sampling on image manifolds using diffusion models
L. Luzi, P. M. Mayer, J. Casco-Rodriguez, A. Siahkoohi, and R. G. Baraniuk
Transactions on Machine Learning Research, 2024.
[pdf] [code] [link] [bib]
Learned multiphysics inversion with differentiable programming and machine learning
M. Louboutin, Z. Yin, R. Orozco, T. J. Grady II, A. Siahkoohi, G. Rizzuti, P. A. Witte, O. Møyner, G. J. Gorman, and F. J. Herrmann
The Leading Edge, 42(7):474–486, 2023.
[pdf] [link] [bib] [featured in Seismic Soundoff]
Optimized time-lapse acquisition design via spectral gap ratio minimization
Y. Zhang, Z. Yin, O. López, A. Siahkoohi, M. Louboutin, R. Kumar, and F. J. Herrmann
Geophysics, 88(4):A19–A23, 2023.
[pdf] [link] [bib]
Reliable amortized variational inference with physics-based latent distribution correction
A. Siahkoohi, G. Rizzuti, R. Orozco, and F. J. Herrmann
Geophysics, 88(3):R297–R322, 2023.
[pdf] [slides] [code] [link] [bib] [featured in Geophysics Bright Spots]
Deep Bayesian inference for seismic imaging with tasks
A. Siahkoohi, G. Rizzuti, and F. J. Herrmann
Geophysics, 87(5):S281–S302, 2022.
[pdf] [code] [link] [bib]
The importance of transfer learning in seismic modeling and imaging
A. Siahkoohi, M. Louboutin, and F. J. Herrmann
Geophysics, 84(6):A47–A52, 2019.
[pdf] [code] [link] [bib]
Conference proceedings and presentations
Self-consuming generative models go MAD
S. Alemohammad, J. Casco-Rodriguez, L. Luzi, A. I. Humayun, H. Babaei, D. LeJeune, A. Siahkoohi, and R. G. Baraniuk
The Twelfth International Conference on Learning Representations, 2024.
[pdf] [extended pdf] [poster] [link] [bib] [featured in the news 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12]
Titan: Bringing the deep image prior to implicit representations
L. Luzi, D. LeJeune, A. Siahkoohi, S. Alemohammad, V. Saragadam, H. Babaei, N. Liu, Z. Wang, and R. G. Baraniuk
IEEE International Conference on Acoustics, Speech and Signal Processing, pages 6165–6169, 2024.
[pdf] [code] [link] [bib]
Conditional score-based diffusion models for Bayesian inference in infinite dimensions
L. Baldassari, A. Siahkoohi, J. Garnier, K. Sølna, and M. V. de Hoop
Advances in Neural Information Processing Systems, volume 36, pages 24262–24290, 2023.
[pdf] [slides] [poster] [code] [link] [bib] [featured as a Spotlight presentation]
Unearthing InSights into Mars: Unsupervised source separation with limited data
A. Siahkoohi, R. Morel, M. V. de Hoop, E. Allys, G. Sainton, and T. Kawamura
Proceedings of the 40th International Conference on Machine Learning, volume 202, pages 31754–31772, 2023.
[pdf] [slides] [poster] [code] [link] [bib]
Amortized normalizing flows for transcranial ultrasound with uncertainty quantification
R. Orozco, M. Louboutin, A. Siahkoohi, G. Rizzuti, T. van Leeuwen, and F. J. Herrmann
Medical Imaging with Deep Learning, volume 227, pages 332–349, 2023.
[pdf] [link] [bib]
Refining amortized posterior approximations using gradient-based summary statistics
R. Orozco, A. Siahkoohi, M. Louboutin, and F. J. Herrmann
5th Symposium on Advances in Approximate Bayesian Inference, 2023.
[pdf] [link] [bib]
Adjoint operators enable fast and amortized machine learning based Bayesian uncertainty quantification
R. Orozco, A. Siahkoohi, G. Rizzuti, T. van Leeuwen, and F. J. Herrmann
Medical Imaging 2023: Image Processing, volume 12464, page 124641L, 2023.
[pdf] [link] [bib]
Learned one-shot imaging
M. Louboutin, R. Orozco, A. Siahkoohi, and F. J. Herrmann
Third International Meeting for Applied Geoscience & Energy, 2023.
[html] [poster] [code] [bib]
3D seismic survey design by maximizing the spectral gap
Y. Zhang, Z. Yin, O. Lopez, A. Siahkoohi, M. Louboutin, and F. J. Herrmann
Third International Meeting for Applied Geoscience & Energy, 2023.
[pdf] [poster] [bib]
Ultra-low-bitrate speech coding with pretrained Transformers
A. Siahkoohi, M. Chinen, T. Denton, W. B. Kleijn, and J. Skoglund
Proceedings of Interspeech, pages 4421–4425, 2022.
[pdf] [link] [bib]
Velocity continuation with Fourier neural operators for accelerated uncertainty quantification
A. Siahkoohi, M. Louboutin, and F. J. Herrmann
Society of Exploration Geophysicists Technical Program Expanded Abstracts, pages 1765–1769, 2022.
[pdf] [slides] [code] [link] [bib]
Accelerating innovation with software abstractions for scalable computational geophysics
M. Louboutin, P. Witte, A. Siahkoohi, G. Rizzuti, Z. Yin, R. Orozco, and F. J. Herrmann
Society of Exploration Geophysicists Technical Program Expanded Abstracts, pages 1482–1486, 2022.
[pdf] [slides] [link] [bib]
Learned coupled inversion for carbon sequestration monitoring and forecasting with Fourier neural operators
Z. Yin, A. Siahkoohi, M. Louboutin, and F. J. Herrmann
Society of Exploration Geophysicists Technical Program Expanded Abstracts, pages 467–472, 2022.
[pdf] [slides] [code] [link] [bib]
A simulation-free seismic survey design by maximizing the spectral gap
Y. Zhang, M. Louboutin, A. Siahkoohi, Z. Yin, R. Kumar, and F. J. Herrmann
Society of Exploration Geophysicists Technical Program Expanded Abstracts, pages 15–20, 2022.
[pdf] [slides] [code] [link] [bib]
Wave-equation based inversion with amortized variational bayesian inference
A. Siahkoohi, R. Orozco, G. Rizzuti, and F. J. Herrmann
EAGE Workshop on Deep Learning for Seismic Processing: Investigating the Foundations, 2022.
[pdf] [slides] [code] [link] [bib]
Photoacoustic imaging with conditional priors from normalizing flows
R. Orozco, A. Siahkoohi, G. Rizzuti, T. van Leeuwen, and F. J. Herrmann
Neural Information Processing Systems Workshop on Deep Learning and Inverse Problems, 2021.
[pdf] [poster] [link] [bib]
Preconditioned training of normalizing flows for variational inference in inverse problems
A. Siahkoohi, G. Rizzuti, M. Louboutin, P. Witte, and F. J. Herrmann
3rd Symposium on Advances in Approximate Bayesian Inference, 2021.
[pdf] [slides] [code] [link] [bib]
Learning by example: Fast reliability-aware seismic imaging with normalizing flows
A. Siahkoohi and F. J. Herrmann
Society of Exploration Geophysicists Technical Program Expanded Abstracts, pages 1580–1585, 2021.
[pdf] [slides] [code] [link] [bib]
Enabling uncertainty quantification for seismic data preprocessing using normalizing flows (NF)—An interpolation example
R. Kumar, M. Kotsi, A. Siahkoohi, and A. Malcolm
Society of Exploration Geophysicists Technical Program Expanded Abstracts, pages 1515–1519, 2021.
[pdf] [code] [link] [bib]
Parameterizing uncertainty by deep invertible networks, an application to reservoir characterization
G. Rizzuti, A. Siahkoohi, P. A. Witte, and F. J. Herrmann
Society of Exploration Geophysicists Technical Program Expanded Abstracts, pages 1541–1545, 2020.
[pdf] [slides] [code] [link] [bib]
Transfer learning in large-scale ocean bottom seismic wavefield reconstruction
M. Zhang, A. Siahkoohi, and F. J. Herrmann
Society of Exploration Geophysicists Technical Program Expanded Abstracts, pages 1666–1670, 2020.
[pdf] [slides] [code] [link] [bib]
Weak deep priors for seismic imaging
A. Siahkoohi, G. Rizzuti, and F. J. Herrmann
Society of Exploration Geophysicists Technical Program Expanded Abstracts, pages 2998–3002, 2020.
[pdf] [slides] [code] [link] [bib]
Uncertainty quantification in imaging and automatic horizon tracking—a Bayesian deep-prior based approach
A. Siahkoohi, G. Rizzuti, and F. J. Herrmann
Society of Exploration Geophysicists Technical Program Expanded Abstracts, pages 1636–1640, 2020.
[pdf] [slides] [code] [link] [bib]
A deep-learning based Bayesian approach to seismic imaging and uncertainty quantification
A. Siahkoohi, G. Rizzuti, and F. J. Herrmann
European Association of Geoscientists & Engineers Conference and Exhibition Extended Abstracts, 2020.
[pdf] [slides] [code] [link] [bib]
Learned imaging with constraints and uncertainty quantification
F. J. Herrmann, A. Siahkoohi, and G. Rizzuti
Neural Information Processing Systems Deep Inverse Workshop, 2019.
[pdf] [slides] [poster] [link] [bib]
Deep-learning based ocean bottom seismic wavefield recovery
A. Siahkoohi, R. Kumar, and F. J. Herrmann
Society of Exploration Geophysicists Technical Program Expanded Abstracts, pages 2232–2237, 2019.
[pdf] [TensorFlow code] [PyTorch code] [slides] [link] [bib]
Surface-related multiple elimination with deep learning
A. Siahkoohi, D. J. Verschuur, and F. J. Herrmann
Society of Exploration Geophysicists Technical Program Expanded Abstracts, pages 4629–4634, 2019.
[pdf] [slides] [link] [bib]
Learned iterative solvers for the Helmholtz equation
G. Rizzuti, A. Siahkoohi, and F. J. Herrmann
European Association of Geoscientists & Engineers Conference and Exhibition Extended Abstracts, 2019.
[pdf] [slides] [link] [bib]
Deep convolutional neural networks in prestack seismic—two exploratory examples
A. Siahkoohi, M. Louboutin, R. Kumar, and F. J. Herrmann
Society of Exploration Geophysicists Technical Program Expanded Abstracts, pages 2196–2200, 2018.
[pdf] [poster] [link] [bib]
Seismic data reconstruction with generative adversarial networks
A. Siahkoohi, R. Kumar, and F. J. Herrmann
European Association of Geoscientists & Engineers Conference and Exhibition Extended Abstracts, 2018.
[pdf] [slides] [link] [bib]
Sparsity promoting least squares migration for laterally inhomogeneous media
A. Siahkoohi and A. Gholami
7th EAGE Saint Petersburg International Conference and Exhibition, 2016.
[pdf] [link] [bib]
A novel utilization of wireless sensor networks as data acquisition system in smart grids
M. S. Ebrahimi, M. H. Daraei, J. Rezaei, and A. Siahkoohi
International Conference on Materials Science and Information Technology, volume 433-440, pages 6725–6730, 2012.
[link] [link] [bib]
A content-based digital image watermarking algorithm robust against JPEG compression
A. Najafi, A. Siahkoohi, and M. B. Shamsollahi
IEEE International Symposium on Signal Processing and Information Technology, pages 432–437, 2011.
[pdf] [link] [bib]
Thesis
Deep generative models for solving geophysical inverse problems
A. Siahkoohi
Ph.D. thesis, Georgia Institute of Technology, 2022.
[pdf] [slides] [link] [bib]
Technical reports
Taming score-based diffusion priors for infinite-dimensional nonlinear inverse problems
L. Baldassari, A. Siahkoohi, J. Garnier, K. Sølna, and M. V. de Hoop
2024.
[pdf] [bib]
Low-memory stochastic backpropagation with multi-channel randomized trace estimation
M. Louboutin, A. Siahkoohi, R. Wang, and F. J. Herrmann
Technical Report arXiv:2106.06998, Georgia Institute of Technology, 2021.
[pdf] [code] [link] [bib]
Faster uncertainty quantification for inverse problems with conditional normalizing flows
A. Siahkoohi, G. Rizzuti, P. A. Witte, and F. J. Herrmann
Technical Report arXiv:2007.07985, Georgia Institute of Technology, 2020.
[pdf] [link] [bib]
Neural network augmented wave-equation simulation
A. Siahkoohi, M. Louboutin, and F. J. Herrmann
Technical Report arXiv:1910.00925, Georgia Institute of Technology, 2019.
[pdf] [code] [link] [bib]
Miscellaneous
“I wish you hadn’t talked so much, it was distracting.” — Patrick Winston, How to Speak
“No one cares about the inside of your head.” — Larry McEnerney, The Craft of Writing Effectively