L dwarfs detection from SDSS images using improved Faster R-CNN

Abstract

We present a data-driven approach to automatically detect L dwarfs from Sloan Digital Sky Survey(SDSS) images using an improved Faster R-CNN framework based on deep learning. The established L dwarf automatic detection (LDAD) model distinguishes L dwarfs from other celestial objects and backgrounds in SDSS field images by learning the features of 387 SDSS images containing L dwarfs. Applying the LDAD model to the SDSS images containing 93 labeled L dwarfs in the test set, we successfully detected 83 known L dwarfs with a recall rate of 89.25% for known L dwarfs. Several techniques are implemented in the LDAD model to improve its detection performance for L dwarfs,including the deep residual network and the feature pyramid network. As a result, the LDAD model outperforms the model of the original Faster R-CNN, whose recall rate of known L dwarfs is 80.65% for the same test set. The LDAD model was applied to detect L dwarfs from a larger validation set including 843 labeled L dwarfs, resulting in a recall rate of 94.42% for known L dwarfs. The newly identified candidates include L dwarfs, late M and T dwarfs, which were estimated from color (i-z) and spectral type relation. The contamination rates for the test candidates and validation candidates are 8.60% and 9.27%, respectively. The detection results indicate that our model is effective to search for L dwarfs from astronomical images.

12 pages, 10 figures, accepted to be published in AJ

References in corpus (20)

• The NumPy array: a structure for efficient numerical computation
• The UKIRT Infrared Deep Sky Survey (UKIDSS)
• Constraining the Age-Activity Relation for Cool Stars: The SDSS DR5 Low-Mass Star Spectroscopic Sample
• Star-galaxy Classification Using Deep Convolutional Neural Networks
• SpeX Spectroscopy of Unresolved Very Low Mass Binaries. II. Identification of Fourteen Candidate Binaries with Late-M/Early-L and T Dwarf Components
• BOSS Ultracool Dwarfs I: Colors and Magnetic Activity of M and L dwarfs
• A Search for L/T Transition Dwarfs With Pan-STARRS1 and WISE. II. L/T Transition Atmospheres and Young Discoveries
• A large spectroscopic sample of L and T dwarfs from UKIDSS LAS: peculiar objects, binaries, and space density
• Wide, Cool and Ultracool Companions to Nearby Stars from Pan-STARRS1
• Primeval very low-mass stars and brown dwarfs. I. Six new L subdwarfs, classification and atmospheric properties
• An Infrared High Proper Motion Survey Using 2MASS and SDSS: Discovery of M, L and T Dwarfs
• Ultra-cool dwarfs: new discoveries, proper motions, and improved spectral typing from SDSS and 2MASS photometric colors
• Photometric brown-dwarf classification. I. A method to identify and accurately classify large samples of brown dwarfs without spectroscopy
• A Brown Dwarf Census from the SIMP Survey
• New ultracool subdwarfs identified in large-scale surveys using Virtual Observatory tools: II. SDSS DR7 vs UKIDSS LAS DR6, SDSS DR7 vs UKIDSS LAS DR8, SDSS DR9 vs UKIDSS LAS DR10, and SDSS DR7 vs 2MASS
• A Statistical Survey of Peculiar L and T Dwarfs in SDSS, 2MASS, and WISE
• Extremely faint high proper motion objects from SDSS stripe 82 - Optical classification spectroscopy of about 40 new objects
• Automatic detection of low surface brightness galaxies from SDSS images
• The Late-Type Extension to MoVeRS (LaTE-MoVeRS): Proper motion verified low-mass stars and brown dwarfs from SDSS, 2MASS, and WISE
• Ultracool Dwarfs in deep extragalactic surveys using the Virtual Observatory: ALHAMBRA and COSMOS