BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:From Optimal Gridding toward Learning-Based Imaging in Radio Inter
 ferometry - Dr Haoyang Ye\, Astrophysics\, Cavendish Laboratory
DTSTART:20260204T093000Z
DTEND:20260204T100000Z
UID:TALK244108@talks.cam.ac.uk
CONTACT:Leona Hope-Coles
DESCRIPTION:Radio interferometric imaging underpins much of modern radio a
 stronomy\, yet its computational cost increasingly limits scientific retur
 n for current facilities and poses a major challenge for the Square Kilome
 tre Array (SKA)\, which will be the largest radio interferometric array ev
 er constructed. A dominant bottleneck lies in the gridding and degridding 
 operations that enable FFT-based imaging from irregularly sampled visibili
 ties\, particularly for wide-field\, high-resolution observations. \n\nIn 
 this talk\, I present my contributions to improving the accuracy and effic
 iency of interferometric imaging\, spanning both optimal gridding methods 
 and wide-field imaging algorithms. My work on gridding focuses on directly
  minimising image-domain error rather than relying on analytic kernel assu
 mptions\, while my wide-field imaging developments reduce the computationa
 l cost of imaging large sky areas at high resolution. These methods have b
 een adopted in widely used imaging software and have enabled high-fidelity
 \, low-frequency\, wide-field images previously considered impractical\, i
 ncluding benchmark LOFAR results at arcsecond resolution. \n\nBuilding on 
 this foundation\, I discuss ongoing and future work aimed at fundamentally
  reducing imaging cost through AI-driven approaches. By combining numerica
 l optimisation\, compact analytic representations\, and machine-learning
 –based amortisation\, this research seeks to deliver imaging methods tha
 t are both scientifically optimal and computationally scalable. More ambit
 iously\, neural operators offer the prospect of bypassing gridding entirel
 y\, learning the visibility-to-image mapping directly from irregularly-sam
 pled data without intermediate gridding\, FFT\, or correction steps.\n
LOCATION:Seminar Room C\, RDC\, Cavendish Laboratory
END:VEVENT
END:VCALENDAR