Journal-first submissions deadline
8 Aug
Priority submissions deadline 30 Jul
Final abstract submissions deadline 15 Oct
Manuscripts due for FastTrack publication
17 Nov

Early registration ends 15 Dec

Short Courses
16 Jan
Symposium begins
17 Jan
All proceedings manuscripts due
31 Jan

Computational Imaging XX

Conference keywords: inverse problems, image reconstruction, image analysis, denoising, model-based imaging

On this page

ATTENTION: EI 2022 will occur live ONLINE. 
The program will include a variety of opportunities to interact with colleagues and presenters live, in formal and informal settings. IS&T is committed to providing the best online experience possible and supporting the community in light of travel/reimbursement restrictions. Join us for EI 2022 . . . submit your work today!

Conference Overview

More than ever before, computers and computation are critical to the image formation process. Across diverse applications and fields, remarkably similar imaging problems appear, requiring sophisticated mathematical, statistical, and algorithmic tools. This conference focuses on imaging as a marriage of computation with physical devices. It emphasizes the interplay between mathematical theory, physical models, and computational algorithms that enable effective current and future imaging systems. Contributions to the conference are solicited on topics ranging from fundamental theoretical advances to detailed system-level implementations and case studies.

2022 Conference Topics

Algorithms and methodologies

  • Inverse methods
  • Model-based imaging and compressed sensing
  • Estimation techniques
  • Imaging system modeling and simulation
  • Optimization approaches
  • Multiscale image processing and modeling
  • Statistical learning and analysis methods

Key problem areas

  • Image recovery from sensor data
  • Denoising, demosaicing, color correction
  • Deblurring and high-resolution rendering
  • Image and color transforms and analysis
  • Visual perception as an inverse problem
  • Tomography, transmission, and emission
  • Microscopy, light, EM, and non-classical
  • Optical coherence imaging
  • MRI, anatomical, functional, and molecular
  • Acoustic imaging
  • Diffusion optical imaging
  • Electrical resistance and impedance imaging
  • Crystallography
  • Synthetic aperture radar
  • Holographic and coherent optical imaging
  • Computational depth-of-field enhancement
  • Intelligent image cropping and scaling
  • Plenoptics and non-classical image capture
  • Coded aperture and compressed sensing

Current and future applications

  • Consumer imaging and computational photography
  • Super-resolution and enhancement
  • Imaging and camera networks
  • Non-destructive evaluation for additive manufacturing
  • Medical imaging and image-guided surgery
  • Microscopy and clinical applications
  • Emerging biomedical applications
  • Geophysical imaging
  • Materials imaging and characterization
  • Nondestructive testing and evaluation
  • Imagery-based surveillance and tracking
  • Target classification and identification
  • Remote sensing applications

2022 Special Session

Photon-limited Imaging Sensors and Algorithms

This session is jointly sponsored by: Computational Imaging XX, and Imaging Sensors and Systems 2022.

Session Organizing Chairs:
Stanley Chan, Purdue University (United States)
Keigo Hirakawa, The University of Dayton (United States)

The proliferation of miniature sensors on mobile devices and robots has significantly pushed the demand for better sensitivity in photon-limited environments. Unconventional sensor designs such as single-photon detectors have begun to gain acceptance in more mainstream applications as we seek the next generation of image sensors. However, the operating principles of these sensors are sometimes very different from the traditional image sensors, in terms of bit-depth, dynamic range, signal response, color, and image processing. This special session brings together experts from sensors, image processing, computer vision, and machine learning to discuss the latest technologies in the field and the open problems we are facing today.

Recent Advances in Scientific CT

Session Organizing Chairs:
Singanallur V Venkatakrishnan, Oak Ridge National Laboratory (United States)
Doga Gursoy, Argonne National Laboratory (United States)
K. Aditya Mohan, Lawrence Livermore National Laboratory (United States)

Scientific computed tomography (CT) is carried out using a variety of sources (electrons, X-rays, protons, neutrons etc.) in order to investigate samples across different length scales ranging from the angstrom to the millimeter scale. Scientific CT instruments are critical to improve understanding of samples relevant to materials sciences, biology etc. and have even been adapted for time-resolved studies. This session will bring together researchers in algorithm development and the instrument sciences to discuss the latest advances in pushing the capabilities of scientific CT instruments including but not limited to - design of novel algorithms to improve resolution, reduction of experiment time especially in the context of shared facilities, and development of easy to use software libraries.

Computational Chemical Imaging

Session Organizing Chairs:
Ji-Xin Cheng, Boston University (United States)
Garth J. Simpson, Purdue University (United States)

The field of chemical imaging has been historically driven by advanced instrumentation and molecular probe development. However, new opportunities are emerging through the integration of computational imaging concepts into the data acquisition and analysis pipelines for chemical-specific microscopy. This symposium highlights recent progress in advancing imaging speed, SNR, and spectral bandwidth through a synergistic integration of advanced instrumentation, machine learning, and computational imaging algorithm development.

Topics in Coherent Sensing

Session Organizing Chairs:
Andre Van Rynbach, Air Force Research Laboratory, Sensors Directorate, Wright-Patterson Air Force Base (United States)

Measurement of the full complex phase and amplitude of a light wave in a coherent imaging system allows for the maximum versatility when reconstructing an image. Once the coherent optical field is characterized, powerful digital reconstruction techniques aided by machine learning can be applied to key problem areas in computational imaging, such as remote sensing, microscopy, digital holography, super resolution, image enhancement, turbulence mitigation, and synthetic aperture radar/lidar. This special session will highlight recent developments in coherent sensing techniques along with the exciting implications for advances in computational imaging.

Latent Fields in Additive Manufacturing: from Sensing to Reconstruction

Session Organizing Chairs:
Jeff Simmons, Air Force Research Laboratory, Wright-Patterson Air Force Base (United States)
Megna Shah, Air Force Research Laboratory, Wright-Patterson Air Force Base (United States)
Amir Ziabari, Oak Ridge National Laboratory (United States)

Additive Manufacturing (AM) is, arguably, the most instrumented materials process ever devised. The overarching goal of printing defect-free and consistent AM parts has required integration of a multitude of sensors during the manufacturing process, as well as during inspection for testing, qualification, and certification of the manufactured parts. To that end in-situ monitoring sensors are deployed to measure temperatures, light and acoustic signals emitted, current consumed, warpage and displacement, as well as many others. Imaging modalities such as X-ray or Ultrasound CT systems or structured light are used for inspection of the manufactured parts. This rich sensor data allows for the latent fields used in physics-based models to be, for the first time, estimated with modern inversion methods. Latent fields such as temperature distributions, composition fields, or stress and strain fields lie at the heart of models used in Materials Science for prediction of properties and performance. An independent inference of these latent fields from the sensor data will allow for almost direct control over the structures formed during processing, a major step towards a “properties on demand” solution. This special session brings together experts in the fields of Materials Science and Signal Processing who are involved in additive manufacturing to explore the possibilities in this emerging field.

Autonomous Science

Session Organizing Chairs:
Doga Gursoy, Argonne National Laboratory (United States)
Benji Maruyama, Air Force Research Laboratory, Materials & Manufacturing Directorate, Wright-Patterson Air Force Base (United States)

We are living on the cusp of an age in which scientific inquiry has moved beyond the comprehensibility of humans. To accelerate the rate of scientific discovery, researchers need new ways to perform scientific experiments that go beyond mere manual skills. A promising route lies in an emerging field known as “Autonomous Science”, where intelligent algorithms design and run experiments and decide themselves on the next steps to take. This type of autonomous capability offers to overcome the challenges faced by computational and experimental facilities worldwide, such as the steady increase in volume and complexity of the scientific data. This special session aims to bring together researchers from relevant disciplines to disseminate information on the state-of-the-art methods and techniques and to enhance community around this field.

2022 Committee

Conference Chairs

Charles A. Bouman, Purdue University (United States)
Gregery T. Buzzard, Purdue University (United States)
Robert L. Stevenson, University of Notre Dame (United States)

Program Committee

Clem Karl, Boston University (United States)
Eric Miller, Tufts University (United States)
Joseph A O'Sullivan, Washington University in St. Louis (United States)
Hector J Santos-Villalobos, Oak Ridge National Laboratory (United States)
Ken D. Sauer, University of Notre Dame (United States)

Community Chair

Begum Gulsoy, Northwestern University (United States)

No content found

No content found

No content found