Searching for pulsars using image pattern recognition

Abstract

In this paper, we present a novel artificial intelligence (AI) program that identifies pulsars from recent surveys using image pattern recognition with deep neural nets---the PICS (Pulsar Image-based Classification System) AI. The AI mimics human experts and distinguishes pulsars from noise and interferences by looking for patterns from candidate. The information from each pulsar candidate is synthesized in four diagnostic plots, which consist of up to thousands pixel of image data. The AI takes these data from each candidate as its input and uses thousands of such candidates to train its ~9000 neurons. Different from other pulsar selection programs which use pre-designed patterns, the PICS AI teaches itself the salient features of different pulsars from a set of human-labeled candidates through machine learning. The deep neural networks in this AI system grant it superior ability in recognizing various types of pulsars as well as their harmonic signals. The trained AI's performance has been validated with a large set of candidates different from the training set. In this completely independent test, PICS ranked 264 out of 277 pulsar-related candidates, including all 56 previously known pulsars, to the top 961 (1%) of 90008 test candidates, missing only 13 harmonics. The first non-pulsar candidate appears at rank 187, following 45 pulsars and 141 harmonics. In other words, 100% of the pulsars were ranked in the top 1% of all candidates, while 80% were ranked higher than any noise or interference. The performance of this system can be improved over time as more training data are accumulated. This AI system has been integrated into the PALFA survey pipeline and has discovered six new pulsars to date.

29 pages, 9 figures, two tables, accepted by ApJ

References in corpus (4)

• A bright millisecond radio burst of extragalactic origin
• A Population of Fast Radio Bursts at Cosmological Distances
• Discovery of 28 pulsars using new techniques for sorting pulsar candidates
• PEACE: Pulsar Evaluation Algorithm for Candidate Extraction -- A software package for post-analysis processing of pulsar survey candidates

Cited by in corpus (64)

• Considerations for a Multi-beam Multi-purpose Survey with FAST
• Fifty Years of Pulsar Candidate Selection: From simple filters to a new principled real-time classification approach
• The Green Bank Northern Celestial Cap Pulsar Survey - I: Survey Description, Data Analysis, and Initial Results
• The FAST Galactic Plane Pulsar Snapshot survey: I. Project design and pulsar discoveries
• Fast Radio Burst 121102 Pulse Detection and Periodicity: A Machine Learning Approach
• Convolutional neural networks: a magic bullet for gravitational-wave detection?
• FETCH: A deep-learning based classifier for fast transient classification
• Arecibo Pulsar Survey Using ALFA. IV. Mock Spectrometer Data Analysis, Survey Sensitivity, and the Discovery of 41 Pulsars
• A Hybrid Ensemble method for Pulsar Candidate Classification
• SPINN: a straightforward machine learning solution to the pulsar candidate selection problem
• The Green Bank North Celestial Cap Pulsar Survey III: 45 New Pulsar Timing Solutions
• Applying Deep Learning to Fast Radio Burst Classification
• Astronomia ex machina: a history, primer, and outlook on neural networks in astronomy
• Machine-learning Approaches to Exoplanet Transit Detection and Candidate Validation in Wide-field Ground-based Surveys
• Separation of pulsar signals from noise with supervised machine learning algorithms
• The MPIfR-MeerKAT Galactic Plane survey I -- System setup and early results
• Deep residual detection of radio frequency interference for FAST
• PALFA Single-Pulse Pipeline: New Pulsars, Rotating Radio Transients and a Candidate Fast Radio Burst
• PSR J1926-0652: A Pulsar with Interesting Emission Properties Discovered at FAST
• Detection of Dispersed Radio Pulses: A machine learning approach to candidate identification and classification
• A Fast Radio Burst discovered in FAST drift scan survey
• FAST discovery of an extremely radio-faint millisecond pulsar from the Fermi-LAT unassociated source 3FGL J0318.1+0252
• An investigation of pulsar searching techniques with the Fast Folding Algorithm
• The TRAPUM L-band survey for pulsars in Fermi-LAT gamma-ray sources
• Einstein@Home Discovery of a PALFA Millisecond Pulsar in an Eccentric Binary Orbit
• Concat Convolutional Neural Network for Pulsar Candidate Selection
• Pulsar Candidate Identification Using Semi-Supervised Generative Adversarial Networks
• The implementation of a Fast-Folding pipeline for long-period pulsar searching in the PALFA survey
• Multivariate Classification with Random Forests for Gravitational Wave Searches of Black Hole Binary Coalescence
• Pulsar Candidates Classification with Deep Convolutional Neural Networks
• Study of 72 pulsars discovered in the PALFA survey: Timing analysis, glitch activity, emission variability, and a pulsar in an eccentric binary
• Pulsars Detection by Machine Learning with Very Few Features
• Timing of Five PALFA-Discovered Millisecond Pulsars
• Classification of Planetary Nebulae through Deep Transfer Learning
• PulsarX: a new pulsar searching package -I. A high performance folding program for pulsar surveys
• Connecting optical morphology, environment, and HI mass fraction for low-redshift galaxies using deep learning
• Missing for 20 years: MeerKAT re-detects the elusive binary pulsar M30B
• Multi-epoch searches for relativistic binary pulsars and fast transients in the Galactic Centre
• The discovery of a rotating radio transient J1918$-$0449 with intriguing emission properties with the five hundred meter aperture spherical radio telescope
• Pulsar Candidate Sifting Using Multi-input Convolution Neural Networks
• The TRAPUM Small Magellanic Cloud pulsar survey with MeerKAT: I. Discovery of seven new pulsars and two Pulsar Wind Nebula associations
• Pushchino multibeams pulsar search. IV. Detection of new pulsars at declinations -9o < δ< +55o
• TRAPUM search for pulsars in supernova remnants and pulsar wind nebulae -- I. Survey description and initial discoveries
• The TRAPUM Large Magellanic Cloud pulsar survey with MeerKAT I: Survey setup and first seven pulsar discoveries
• A GPU based single-pulse search pipeline (GSP) with database and its application to the commensal radio astronomy FAST survey (CRAFTS)
• Search for fast radio transients using Arecibo drift-scan observations at 1.4 GHz
• Vela Pulsar: Single Pulses Analysis with Machine Learning Techniques
• The SPAN512 mid-latitude pulsar survey at the Nançay Radio Telescope
• Simulating high-time resolution radio-telescope observations
• Machine Learning Pipeline for Pulsar Star Dataset
• TRAPUM upper limits on pulsed radio emission for SMC X-ray pulsar J0058-7218
• TRAPUM pulsar and transient search in the Sextans A and B galaxies and discovery of background FRB 20210924D
• Probabilistic learning for pulsar classification
• Detecting Pulsars with Neural Networks: A Proof of Concept
• A simulation experiment of a pipeline based on machine learning for neutral hydrogen intensity mapping surveys
• Searching for Pulsars, Magnetars, and Fast Radio Bursts in the Sculptor Galaxy using MeerKAT
• Dealing with the data imbalance problem on pulsar candidates sifting based on feature selection
• A Method for Pulsar Searching: Combining a Two-dimensional Autocorrelation Profile Map and a Deep Convolutional Neural Network
• A Processing Pipeline for High Volume Pulsar Data Streams
• QPOML: A Machine Learning Approach to Detect and Characterize Quasi-Periodic Oscillations in X-ray Binaries
• Early-fusion Based Pulsar Identification with Smart Under-sampling
• 200,000+ Deep Learning-inferred Periods of Stellar Variability from the All-Sky Automated Survey for Supernovae
• Radio Galaxy Zoo: Giant Radio Galaxy Classification using Multi-Domain Deep Learning
• Spatial Modulation Search Applied to the Search and Confirmation of Highly Scintillated Pulsars at FAST with A Pulsar Discovered in M3