2nd Workshop on Applications of Computational Imaging (WACI)
4 March, 2025 - Tucson, AZ
https://www.zoomgov.com/j/16072919948?pwd=azVpN0lQekVnS0pSQUVBOXQrRWtjUT09
4 March, 2025 - Tucson, AZ
https://www.zoomgov.com/j/16072919948?pwd=azVpN0lQekVnS0pSQUVBOXQrRWtjUT09
The second Workshop on Applications of Computational Imaging (WACI) will be held in conjunction with WACV 2025 in Tucson, AZ during one of the two workshop days.
Compared to traditional cameras, computational imaging increasingly offers better image quality, additional information (e.g., 3D or multiple views), and reductions in size, weight, or power consumption. While these benefits are most commonly enjoyed in consumer photography with smartphone cameras, computational cameras provide similar benefits to applications in the industrial, scientific, and defense domains. The purpose of this workshop is to bring together researchers interested in the applications of computational imaging in these and other domains, similar to how WACV brings together researchers interested in applications of computer vision more broadly.
8:15AM Workshop welcome and schedule overview
8:30 AM Invited Speaker: Trish Veeder, Program Manager, Defense Advanced Research Projects Agency (DARPA)
Abstract: The Defense Advanced Research Projects Agency (DARPA) seeks to explore the potential of computational imaging, machine learning, and artificial intelligence tools and techniques to enhance the accuracy of passive ranging for tactical and civil applications. Trish Veeder will introduce the Computational Imaging Detection and Ranging (CIDAR) Challenge (https://cidar.darpa.mil/), which will award more than $2M in prizes to teams that employ optical data captured through various filters to accurately determine range out to 20km.
9:00 AM Invited Speaker: Florian Willomitzer, Associate Professor, Wyant College of Optical Sciences - University of Arizona
Title: Coherent computational imaging with synthetic waves
Abstract: This talk discusses how diversity in illumination wavelengths can be utilized to circumvent the problem of phase randomization in scattered coherent optical light fields. Our technique probes the scene at two or more closely spaced optical wavelengths and computationally assembles a complex "synthetic field" at a "synthetic wavelength," which is used for further processing. As the synthetic wavelength is the beat wavelength of the two used optical “carrier” wavelengths, it can be picked orders of magnitudes larger, and the computationally assembled synthetic field becomes immune to scatter. Amongst other applications, the introduced method allows for holographic measurements of hidden objects through scattering media, for interferometric measurements of macroscopic objects with rough surfaces, or to shape “synthetic pulses” to perform Synthetic Light-in-Flight measurements. During the talk, different flavors of the technique will be introduced, including a method to retrieve the complex synthetic field in single-shot.
9:30AM Paper session
Fashionability-Enhancing Outfit Image Editing with Conditional Diffusion Models
M-GAID: A Real-World Dataset for Ghosting Artifact Detection and Removal in Mobile Imaging
Segment Anything in Light Fields for Real-Time Applications via Constrained Prompting
10:00AM Coffee break
10:30AM Invited Speaker: David Brady, J.W. and H.M. Goodman Endowed Chair in Optical Sciences, University of Arizona
Title: Cameras measure the wrong thing.
Abstract: Since the dawn of photography, forward models have been built on the assumption that sampled data should be drawn from the irradiance or spectral density of a focal image. Here we show that that the focal power spectral density is an incomplete measure of the optical data cube and we describe measurement strategies for increasing camera information capacity.
11:15AM Invited Speaker: Matthew T. Young, Engineering Fellow, Raytheon
Abstract: Training recognition models is a data-intensive process that traditionally requires a verbose set of data that's often costly to acquire and refine, or impossible to capture in the real world at scale. Synthetic data can be used to augment or replace real data, but the impact on a model incorporating synthetic data for image-based systems is poorly characterized, and the ultimate impact on real-world performance is not well studied or understood. We demonstrate that real-world performance of a model trained from synthetic data can be reliable and performant. We outline key characteristics of synthetic data that must be met to ensure viable data is generated to create a useful training set, and establish the requirements for a physics-based synthetic generator to be useful for training.
11:45AM Workshop conclusion and related announcements
Researchers working in computational imaging are encouraged to submit papers to the Workshop on Applications of Computational Imaging, to be held in conjunction with the IEEE Winter conference on the Applications of Computer Vision (WACV).
The recent development of computational cameras has led to a revolution in consumer photography, and also has the potential to revolutionize a wide range of applications which rely on the acquisition of visual information. The objective of this workshop is to bring together and inform practitioners of computational imaging about applications of computational cameras across multiple domains including scientific, industrial, and defense applications. We welcome high quality submissions in:
Computational image acquisition
Computational illumination and displays
Advanced image and video processing
Computational optics
Imaging by compressive sensing
Integration of image sensing with machine learning
Organizing and exploiting photo/video data
Capture of appearance, shape, illumination and underlying physical properties
Imaging, display, and Illumination for user interfaces
Physics-based rendering
Mobile Imaging
Workshop papers will be reviewed according to criteria aligned with WACV’s applications track. That means that the paper should demonstrate that the computational imaging technique is well-suited to the application, and that evaluations are consistent with the intended application; a lack of algorithmic novelty, per se, will not be sufficient criteria for rejection.
Accepted papers will be included in the Proceedings of IEEE WACV 2025 & Workshops and will be sent for inclusion into the IEEE Xplore digital library.
December 20, 2024: Submission deadline
January 3, 2025: Author notification
January 15 (tentative extension) January 10, 2025: Camera ready deadline
8 pages (not including references) and supplementary submission following the main WACV conference template.
The author kit/paper template is provided in Latex format via this overleaf template and this dropbox download. All submissions should use this template.
Scott McCloskey, Assistant Director of Computer Vision, Kitware Inc.
Keigo Hirakawa, Professor of Electrical and Computer Engineering, Intelligent Signal Systems Lab, University of Dayton
Suren Jayasuriya, Assistant Professor in the School of Arts Media and Engineering, Imaging Lyceum, Arizona State University
Connor Hashemi, Kitware
Ripon Saha, Arizona State University