Workshop Description
Inverse problems are ubiquitous in science, medicine, and engineering, and research in this area has produced real-world impact in medical tomography, seismic imaging, computational photography, and other domains. The recent rapid progress in learning-based image generation raises exciting opportunities in inverse problems, and this workshop seeks to gather a diverse set of participants who apply machine learning to inverse problems, from mathematicians and computer scientists to physicists and biologists. This gathering will facilitate new collaborations and will help develop more effective, reliable, and trustworthy learning-based solutions to inverse problems.
Schedule
| Time (CDT) | Event |
|---|---|
| 9:00 | Invited talk 1: Namrata Vaswani |
| 9:30 | Invited talk 2: Stella Yu |
| 10:00 | Break and Discussion 1 |
| 10:30 | Contributed talk 1: “Phase Retrieval via Deep Expectation-Consistent Approximation”, Saurav K Shastri, Philip Schniter |
| 10:45 | Contributed talk 2: “Particle Guidance: non-I.I.D. Diverse Sampling with Diffusion Models”, Gabriele Corso, Yilun Xu, Valentin De Bortoli, Regina Barzilay, Tommi Jaakkola |
| 11:00 | Invited talk 3: Jong Chul Ye |
| 11:30 | Invited talk 4: Ben Poole |
| 12:00 | Lunch break |
| 1:30 | Contributed talk 3: “Space-Time Implicit Neural Representations for Atomic Electron Tomography on Dynamic Samples”, Tiffany Chien, Colin Ophus, Laura Waller |
| 1:45 | Contributed talk 4: “AmbientFlow: Invertible generative models from incomplete, noisy imaging measurements”, Varun A. Kelkar, Rucha Deshpande, Arindam Banerjee, Mark Anastasio |
| 2:00 | Poster session and Discussion 2 |
| 3:00 | Invited talk 5: Eric Price |
| 3:30 | Break |
| 4:00 | Invited talk 6: Ellen Zhong |
| 4:30 | Contributed talk 5: “Quantifying Generative Model Uncertainty in Posterior Sampling Methods for Computational Imaging”, Canberk Ekmekci · Mujdat Cetin |
| 4:45 | Contributed talk 6: “Model-adapted Fourier sampling for generative compressed sensing”, Aaron Berk, Simone Brugiapaglia, Yaniv Plan, Matthew Scott, Xia Sheng, Ozgur Yilmaz |
| 5:00 | End of official program |
Speakers
 |
 |
 |
| Namrata Vaswani Iowa State University |
Ellen Zhong Princeton University |
Jong Chul Ye KAIST |
 |
 |
 |
| Stella Yu University of Michigan, Ann Arbor |
Ben Poole Google Brain |
Eric Price UT Austin |
Organizers
 |
 |
 |
| Christopher Metzler University of Maryland |
Shirin Jalali Rutgers University |
Ajil Jalal UC Berkeley |
 |
 |
 |
| Paul Hand Northeastern University |
Reinhard Heckel Technical University of Munich |
Jon Tamir UT Austin |
 |
 |
|
| Arian Maleki Columbia University |
Richard Baraniuk Rice University |
Call for Papers and Submission Instructions
We invite researchers to submit anonymous papers of up to 4 pages (excluding references and appendices) which will be considered for contributed workshop papers. No specific formatting is required. Authors are encouraged to use the NeurIPS style file, but they may use any other style as long as it has standard font size (11pt) and margins (1in).
Submission at OpenReview is open now until the submission deadline on October 6, 2023.
We welcome all submissions in the intersection of inverse problems and deep learning, including but not limited to submissions on the following topics:
-
Fundamental approaches to address model uncertainty in learning-based solutions for inverse problems: Currently, the best DL-based solutions heavily rely on knowing the inverse system’s forward model and assume simple models of distortion (such as additive Gaussian noise). What algorithms and analysis techniques do we require for applications where we only have access to partial information about the system model?
-
Diffusion models: Diffusion models have recently gained attention as powerful learned priors for solving inverse problems, due to their ability to model complex high-dimensional data across diverse modalities such as MRI, acoustics, graphs, proteins, etc. What are their benefits and limitations, and what are the optimal algorithms?
Important Dates and Logistics
- Submission Deadline: October 6, 2023, 11:59 PM UTC.
- Notification of acceptance: October 20, 2023.
- Workshop: Saturday, December 16, 2023.
- Location: Room 214, Level 2
Dual Submission Policy
This is a non-archival workshop. Hence, you are allowed to resubmit to other venues, as long as the other venue permits it.
If your main paper was accepted at NeurIPS and you wish to submit to this workshop, we expect: (a) the two manuscripts to have sufficient differences in experiments or theory, and (b) an author to physically present accepted posters on the day of the workshop.
Please email deepinverse@gmail.com with any questions.