Fast Holographic Deconvolution: a new technique for precision radio interferometry
Ian Sullivan, Miguel Morales, Bryna Hazelton, Wayne Arcus, David Barnes, Gianni Bernardi, Frank Briggs, Judd D. Bowman, John Bunton, Roger Cappallo, Brian Corey… 展开作者
Ian Sullivan, Miguel Morales, Bryna Hazelton, Wayne Arcus, David Barnes, Gianni Bernardi, Frank Briggs, Judd D. Bowman, John Bunton, Roger Cappallo, Brian Corey, Avinash Deshpande, Ludi deSouza, David Emrich, B. M. Gaensler, Robert Goeke, Lincoln Greenhill, David Herne, Jacqueline Hewitt, Melanie Johnston-Hollitt, David Kaplan, Justin Kasper, Barton Kincaid, Ronald Koenig, Eric Kratzenberg, Colin Lonsdale, Mervyn Lynch, Russell McWhirter, Daniel Mitchell, Edward Morgan, Divya Oberoi, Stephen Ord, Joseph Pathikulangara, Thiagaraj Prabu, Ron Remillard, Alan Rogers, Anish Roshi, Joseph Salah, Robert Sault, Udaya Shankar, K. Srivani, Jamie Stevens, Ravi Subrahmanyan, Steven Tingay, Randall Wayth, Mark Waterson, Rachel Webster, Alan Whitney, Andrew Williams, Chris Williams, Stuart Wyithe
We introduce the Fast Holographic Deconvolution method for analyzing interferometric radio data. Our new method is an extension of A-projection/software-holography/forward modeling analysis techniques and shares their precision deconvolution and widefield polarimetry, while being significantly faster than current implementations that use full direction-dependent antenna gains. Using data from the MWA 32 antenna prototype, we demonstrate the effectiveness and precision of our new algorithm. Fast Holographic Deconvolution may be particularly important for upcoming 21 cm cosmology observations of the Epoch of Reionization and Dark Energy where foreground subtraction is intimately related to the precision of the data reduction.
ShapeNet: Shape Constraint for Galaxy Image Deconvolution
深度学习（DL）在解决各个领域的反问题方面取得了显着的成果。特别是，Tikhonet方法非常强大，可以对光学天文图像进行去卷积（Sureau et al. 2020）。然而，该方法仅使用了$\ell_2$损失，这不能保证图像中重建的对象的物理信息（例如，通量和形状）的保留。在Nammour et al.（2021）中，在稀疏反卷积的框架下提出了一种新的损失函数，该函数更好地保留了星系的形状并降低了像素误差。在本文中，我们扩展Tikhonet考虑到这种形状约束，并应用我们的新DL方法，称为ShapeNet，光学和无线电干涉模拟数据集。本文的原创性依赖于i）我们在神经网络框架中使用的形状约束，ii）深度学习首次应用于无线电干涉图像反卷积，iii）我们为社区提供的模拟无线电数据集的生成。一系列的例子说明了结果。
POLISH'ing the Sky: Wide-Field and High-Dynamic Range Interferometric Image Reconstruction with Application to Strong Lens Discovery
Department of Computing and Mathematical Sciences, California Institute of Technology (Caltech), Pasadena, CA, USA（计算与数学科学系，加州理工学院（Caltech），帕萨迪纳，CA，美国）；Center for Astrophysics ∣ \mid Harvard & Smithsonian, Cambridge, MA, USA（天体物理中心，哈佛与史密松尼天体物理研究中心，剑桥，MA，美国）
