Exoplanet Imaging via Differentiable Rendering

作     者：Feng, Brandon Y. Ferrer-Chavez, Rodrigo Levis, Aviad Wang, Jason J. Bouman, Katherine L. Freeman, William T. 

作者机构：MIT Comp Sci & Artificial Intelligence Lab Cambridge MA 02139 USA Northwestern Univ Dept Phys & Astron Evanston IL 60208 USA Univ Toronto Dept Comp Sci Toronto ON M5S 1A1 Canada CALTECH Dept Comp & Math Sci Pasadena CA 91125 USA CALTECH Dept Elect Engn & Astron Pasadena CA 91125 USA 

出 版 物：《IEEE TRANSACTIONS ON COMPUTATIONAL IMAGING》 (IEEE Trans. Comput. Imaging)

年 卷 期：2025年第11卷

页      面：36-51页

核心收录：

学科分类：0808[工学-电气工程] 1002[医学-临床医学] 08[工学] 

基　　金：NSF NSF AI Institute for Artificial Intelligence and Fundamental Interactions CIF Occlusion and Directional Resolution in Computational Imaging NSERC STScI [JWST-ERS-01386, JWST-GO-04050] NASA [NAS5-03127] NSF CAREER Amazon AI4Science Partnership Discovery Grant Carver Mead New Adventures Fund 

主　　题：Planets Imaging Stars Exoplanet Telescopes Rendering (computer graphics) Extraterrestrial measurements Optical imaging Optimization Optical diffraction Exoplanet imaging high-contrast imaging differentiable rendering wavefront aberration estimation 

摘      要：Direct imaging of exoplanets is crucial for advancing our understanding of planetary systems beyond our solar system, but it faces significant challenges due to the high contrast between host stars and their planets. Wavefront aberrations introduce speckles in the telescope science images, which are patterns of diffracted starlight that can mimic the appearance of planets, complicating the detection of faint exoplanet signals. Traditional post-processing methods, operating primarily in the image intensity domain, do not integrate wavefront sensing data. These data, measured mainly for adaptive optics corrections, have been overlooked as a potential resource for post-processing, partly due to the challenge of the evolving nature of wavefront aberrations. In this paper, we present a differentiable rendering approach that leverages these wavefront sensing data to improve exoplanet detection. Our differentiable renderer models wave-based light propagation through a coronagraphic telescope system, allowing gradient-based optimization to significantly improve starlight subtraction and increase sensitivity to faint exoplanets. Simulation experiments based on the James Webb Space Telescope configuration demonstrate the effectiveness of our approach, achieving substantial improvements in contrast and planet detection limits. Our results showcase how the computational advancements enabled by differentiable rendering can revitalize previously underexploited wavefront data, opening new avenues for enhancing exoplanet imaging and characterization.