Supervised detection of exoplanets in high-contrast imaging sequences

Abstract

Post-processing algorithms play a key role in pushing the detection limits of high-contrast imaging (HCI) instruments. State-of-the-art image processing approaches for HCI enable the production of science-ready images relying on unsupervised learning techniques, such as low-rank approximations, for generating a model PSF and subtracting the residual starlight and speckle noise. In order to maximize the detection rate of HCI instruments and survey campaigns, advanced algorithms with higher sensitivities to faint companions are needed, especially for the speckle-dominated innermost region of the images. We propose a reformulation of the exoplanet detection task (for ADI sequences) that builds on well-established machine learning techniques to take HCI post-processing from an unsupervised to a supervised learning context. In this new framework, we present algorithmic solutions using two different discriminative models: SODIRF (random forests) and SODINN (neural networks). We test these algorithms on real ADI datasets from VLT/NACO and VLT/SPHERE HCI instruments. We then assess their performances by injecting fake companions and using receiver operating characteristic analysis. This is done in comparison with state-of-the-art ADI algorithms, such as ADI-PCA. This study shows the improved sensitivity vs specificity trade-off of the proposed approach. At the diffraction limit, SODINN improves the true positive rate by a factor ranging from ~2 to ~10 (depending on the dataset and angular separation) with respect to ADI-PCA when working at the same false positive level. The proposed supervised detection framework outperforms state-of-the-art techniques in the task of discriminating planet signal from speckles. In addition, it offers the possibility of re-processing existing HCI databases to maximize their scientific return and potentially improve the demographics of directly imaged exoplanets.

Accepted for publication in Astronomy & Astrophysics

References in corpus (12)

• Direct Imaging of Multiple Planets Orbiting the Star HR 8799
• Rotation-invariant convolutional neural networks for galaxy morphology prediction
• Fundamental limitations of high contrast imaging set by small sample statistics
• Generative Adversarial Networks recover features in astrophysical images of galaxies beyond the deconvolution limit
• Star-galaxy Classification Using Deep Convolutional Neural Networks
• VIP: Vortex Image Processing package for high-contrast direct imaging
• Improving and Assessing Planet Sensitivity of the GPI Exoplanet Survey with a Forward Model Matched Filter
• First On-Sky High Contrast Imaging with an Apodizing Phase Plate
• The First Science Results from SPHERE: Disproving the Predicted Brown Dwarf around V471 Tau
• Unsupervised feature-learning for galaxy SEDs with denoising autoencoders
• S4: A Spatial-Spectral model for Speckle Suppression
• Adaptive optics in high-contrast imaging

Cited by in corpus (22)

• MIRACLES: atmospheric characterization of directly imaged planets and substellar companions at 4-5 $\mu$m. I. Photometric analysis of $\beta$ Pic b, HIP 65426 b, PZ Tel B and HD 206893 B
• Surveying the reach and maturity of machine learning and artificial intelligence in astronomy
• Astronomia ex machina: a history, primer, and outlook on neural networks in astronomy
• PynPoint: a modular pipeline architecture for processing and analysis of high-contrast imaging data
• STIM map: detection map for exoplanets imaging beyond asymptotic Gaussian residual speckle noise
• Toward Robust Autotuning of Noisy Quantum Dot Devices
• Using Deep Neural Networks to compute the mass of forming planets
• Comparing Apples with Apples: Robust Detection Limits for Exoplanet High-Contrast Imaging in the Presence of non-Gaussian Noise
• Detrending Exoplanetary Transit Light Curves with Long Short-Term Memory Networks
• RSM detection map for direct exoplanet detection in ADI sequences
• Half-sibling regression meets exoplanet imaging: PSF modeling and subtraction using a flexible, domain knowledge-driven, causal framework
• Impacts of high-contrast image processing on atmospheric retrievals
• Use the 4S (Signal-Safe Speckle Subtraction): Explainable Machine Learning reveals the Giant Exoplanet AF Lep b in High-Contrast Imaging Data from 2011
• Spectral and angular differential imaging with SPHERE/IFS. Assessing the performance of various PCA-based approaches to PSF subtraction
• Pushing the Limits of Exoplanet Discovery via Direct Imaging with Deep Learning
• Recurrence quantification analysis as a post-processing technique in adaptive optics high-contrast imaging
• Auto-RSM: an automated parameter-selection algorithm for the RSM map exoplanet detection algorithm
• NA-SODINN: a deep learning algorithm for exoplanet image detection based on residual noise regimes
• Eliminating artefacts in Polarimetric Images using Deep Learning
• Machine learning for exoplanet detection in high-contrast spectroscopy Combining cross correlation maps and deep learning on medium-resolution integral-field spectra
• Improving the RSM map exoplanet detection algorithm: PSF forward modelling and optimal selection of PSF subtraction techniques
• An Alternating Minimization Algorithm with Trajectory for Direct Exoplanet Detection -- The AMAT Algorithm