Effective Image Differencing with ConvNets for Real-time Transient Hunting

Large sky surveys are increasingly relying on image subtraction pipelines for real-time (and archival) transient detection. In this process one has to contend with varying point-spread function (PSF) and small brightness variations in many sources, as well as artefacts resulting from saturated stars and, in general, matching errors. Very often the differencing is done with a reference image that is deeper than individual images and the attendant difference in noise characteristics can also lead to artefacts. We present here a deep-learning approach to transient detection that encapsulates all the steps of a traditional image-subtraction pipeline – image registration, background subtraction, noise removal, PSF matching and subtraction – in a single real-time convolutional network. Once trained, the method works lightening-fast and, given that it performs multiple steps in one go, the time saved and false positives eliminated for multi-CCD surveys like Zwicky Transient Facility and Large Synoptic Survey Telescope will be immense, as millions of subtractions will be needed per night.

[1]  D. M. Bramich,et al.  A new algorithm for difference image analysis , 2008, 0802.1273.

[2]  W. M. Wood-Vasey,et al.  The Pan-STARRS1 Surveys , 2016, 1612.05560.

[3]  Octavi Fors,et al.  Evryscope Science: Exploring the Potential of All-Sky Gigapixel-Scale Telescopes , 2015, 1501.03162.

[4]  Lin Yan,et al.  The IPAC Image Subtraction and Discovery Pipeline for the Intermediate Palomar Transient Factory , 2016, 1608.01733.

[5]  B. A. Boom,et al.  GW170104: Observation of a 50-Solar-Mass Binary Black Hole Coalescence at Redshift 0.2. , 2017, Physical review letters.

[6]  E. Bertin,et al.  SExtractor: Software for source extraction , 1996 .

[7]  S. Howerton,et al.  CRTS SNHunt: The First Five Years of Supernova Discoveries , 2017 .

[8]  Akshay Pai,et al.  Deep-learnt classification of light curves , 2017, 2017 IEEE Symposium Series on Computational Intelligence (SSCI).

[9]  A. Katsaggelos,et al.  Gravity Spy: integrating advanced LIGO detector characterization, machine learning, and citizen science , 2016, Classical and quantum gravity.

[10]  Thomas Brox,et al.  Hybrid Learning of Optical Flow and Next Frame Prediction to Boost Optical Flow in the Wild , 2017 .

[11]  C. Lintott,et al.  Galaxy Zoo 2: detailed morphological classifications for 304,122 galaxies from the Sloan Digital Sky Survey , 2013, 1308.3496.

[12]  S. G. Djorgovski,et al.  The Palomar-Quest digital synoptic sky survey , 2007, 0801.3005.

[13]  Daan Wierstra,et al.  Stochastic Backpropagation and Approximate Inference in Deep Generative Models , 2014, ICML.

[14]  S. G. Djorgovski,et al.  Towards an Automated Classification of Transient Events in Synoptic Sky Surveys , 2011, CIDU.

[15]  Geoffrey E. Hinton,et al.  Learning representations by back-propagating errors , 1986, Nature.

[16]  Ernest E. Croner,et al.  The Palomar Transient Factory: System Overview, Performance, and First Results , 2009, 0906.5350.

[17]  E. Ofek,et al.  PROPER IMAGE SUBTRACTION—OPTIMAL TRANSIENT DETECTION, PHOTOMETRY, AND HYPOTHESIS TESTING , 2016, 1601.02655.

[18]  N. S. Philip,et al.  Transient Classification in LIGO data using Difference Boosting Neural Network , 2016, 1609.07259.

[19]  Nima Sedaghat.  Next-Flow: Hybrid Multi-Tasking with Next-Frame Prediction to Boost Optical-Flow Estimation in the Wild , 2016, ArXiv.

[20]  Eric C. Bellm,et al.  The Zwicky Transient Facility , 2013, 1410.8185.

[21]  Yann LeCun,et al.  Deep multi-scale video prediction beyond mean square error , 2015, ICLR.

[22]  Christophe Fiorio,et al.  Two Linear Time Union-Find Strategies for Image Processing , 1996, Theor. Comput. Sci..

[23]  Arie Shoshani,et al.  Optimizing connected component labeling algorithms , 2005, SPIE Medical Imaging.

[24]  R. Lupton,et al.  A Method for Optimal Image Subtraction , 1997, astro-ph/9712287.

[25]  John W. Fowler,et al.  Aperture Photometry Tool , 2012 .

[26]  Andrew Zisserman,et al.  Very Deep Convolutional Networks for Large-Scale Image Recognition , 2014, ICLR.

[27]  J. Prieto,et al.  THE MAN BEHIND THE CURTAIN: X-RAYS DRIVE THE UV THROUGH NIR VARIABILITY IN THE 2013 ACTIVE GALACTIC NUCLEUS OUTBURST IN NGC 2617 , 2013, 1310.2241.

[28]  Steve B. Howell,et al.  TWO-DIMENSIONAL APERTURE PHOTOMETRY: SIGNAL-TO-NOISE RATIO OF POINT-SOURCE OBSERVATIONS AND OPTIMAL DATA-EXTRACTION TECHNIQUES , 1989 .

[29]  Tom Charnock,et al.  Deep Recurrent Neural Networks for Supernovae Classification , 2016, ArXiv.

[30]  Pablo A. Estévez,et al.  Deep-HiTS: Rotation Invariant Convolutional Neural Network for Transient Detection , 2017, ArXiv.

[31]  Eduardo Serrano,et al.  LSST: From Science Drivers to Reference Design and Anticipated Data Products , 2008, The Astrophysical Journal.

[32]  Yoshua Bengio,et al.  Gradient-based learning applied to document recognition , 1998, Proc. IEEE.

[33]  Trevor Darrell,et al.  Caffe: Convolutional Architecture for Fast Feature Embedding , 2014, ACM Multimedia.

[34]  A. J. Drake,et al.  FIRST RESULTS FROM THE CATALINA REAL-TIME TRANSIENT SURVEY , 2008, 0809.1394.

[35]  Paul M. Brunet,et al.  The Gaia mission , 2013, 1303.0303.

[36]  Naonori Ueda,et al.  Machine-learning selection of optical transients in the Subaru/Hyper Suprime-Cam survey , 2016, 1609.03249.