A proof-of-concept neural network for inferring parameters of a black hole from partial interferometric images of its shadow

Abstract

We test the possibility of using a convolutional neural network to infer the inclination angle of a black hole directly from the incomplete image of the black hole's shadow in the $uv$-plane. To this end, we develop a proof-of-concept network and use it to explicitly find how the error depends on the degree of coverage, type of input and coverage pattern. We arrive at a typical error of $10^{\circ }$ at a level of absolute coverage $1\%$ (for a pattern covering a central part of the $uv$-plane), $0.3\%$ (pattern covering the central part and the periphery, the $0.3\%$ referring to the central part only), and $14\%$ (uniform pattern). These numbers refer to a network that takes both amplitude and phase of the visibility function as inputs. We find that this type of network works best in terms of the error itself and its distribution for different angles. In addition, the same type of network demonstrates similarly good performance on highly blurred images mimicking sources nearing being unresolved. In terms of coverage, the magnitude of the error does not change much as one goes from the central pattern to the uniform one. We argue that this may be due to the presence of a typical scale which can be mostly learned by the network from the central part alone.

12 pages, 10 figures. For the code and trained models, see https://bitbucket.org/cosmoVlad/neuro-repo

References in corpus (21)

• The NumPy array: a structure for efficient numerical computation
• First M87 Event Horizon Telescope Results. I. The Shadow of the Supermassive Black Hole
• First M87 Event Horizon Telescope Results. VI. The Shadow and Mass of the Central Black Hole
• First M87 Event Horizon Telescope Results. IV. Imaging the Central Supermassive Black Hole
• First M87 Event Horizon Telescope Results. V. Physical Origin of the Asymmetric Ring
• First M87 Event Horizon Telescope Results. II. Array and Instrumentation
• First M87 Event Horizon Telescope Results. III. Data Processing and Calibration
• Event-horizon-scale structure in the supermassive black hole candidate at the Galactic Centre
• Jet Launching Structure Resolved Near the Supermassive Black Hole in M87
• Fast Automated Analysis of Strong Gravitational Lenses with Convolutional Neural Networks
• The Shadow of a Spherically Accreting Black Hole
• Uncertainties in Parameters Estimated with Neural Networks: Application to Strong Gravitational Lensing
• 230 GHz VLBI observations of M87: event-horizon-scale structure at the enhanced very-high-energy $\gamma$-ray state in 2012
• The Size, Shape, and Scattering of Sagittarius A* at 86 GHz: First VLBI with ALMA
• Recurrent Inference Machines for Solving Inverse Problems
• The Chinese space millimeter-wavelength VLBI array - a step toward imaging the most compact astronomical objects
• Asymmetric structure in Sgr A* at 3mm from closure phase measurements with VLBA, GBT and LMT
• Using Deep Neural Networks to compute the mass of forming planets
• Space very long baseline interferometry in China
• An Illustrated History of Black Hole Imaging : Personal Recollections (1972-2002)
• A proof-of-concept neural network for inferring parameters of a black hole from partial interferometric images of its shadow

Cited by in corpus (1)

• A proof-of-concept neural network for inferring parameters of a black hole from partial interferometric images of its shadow