What: A meeting on imaging science in astronomy and beyond, to celebrate the 70th anniversary of Australian astronomer Ron Bracewell’s paper Strip Integration in Radio Astronomy, later a foundation of CT scanning.Where: 17 Wally's Walk, Room 330X, Macquarie UniversityWhen: 8-9 April 2026Why: Find opportunities to have real-world applications for astronomy technology, and for astronomers to benefit from recent advances in other disciplinesJoining: Abstract submissions open soon!
About
Machine learning has revolutionised imaging science, and its impacts are emerging piecemeal in different domains. To adapt William Gibson, “The future is already here — it's just not very evenly distributed.” We want to bring experts across imaging science together to find points of common interest and collaboration. If Bracewell’s innovation in Fourier mathematics was the key that unlocked both radio imaging and CT scanning, today we have a wide-ranging and rapidly-changing set of revolutions from machine learning to understand, advance, and incorporate into our sciences, inSimulation: imaging systems are rooted in the physics of how signals propagate and how they are recorded, and we solve the inverse problem of reconstructing a scene and its salient features from these data. Advances in machine learning mean that we can flexibly connect physical simulations (eg differentiable rendering of the propagation of light) and neural network emulators (eg models for the calibration of detectors or complicated scenes). How should we use and distinguish these for the design, commissioning, and operation of imagers?Representation: multimodal foundation models now exist with extremely strong performance at classification, regression, and generation over images and their metadata. There is more to this than producing AI slop - there are rich representations of images that can be used to express and compress complicated processes from galaxies to internal organs. How can these help the reconstruction or interpretation of imaging?Application: These advances have dramatically extended the capability of imaging systems, so that it barely makes sense to refer to a subfield of “computational imaging” – this is now the core of what we do. What can we use this for in the real world? What are the technologies in one domain that can be usefully applied in another?Abstract submission will be open soon.
Contact
Organizing Committee Chair: Benjamin Pope
Macquarie University[email protected]