scholarly journals ConvoSource: Radio-Astronomical Source-Finding with Convolutional Neural Networks

Galaxies ◽  
2019 ◽  
Vol 8 (1) ◽  
pp. 3
Author(s):  
Vesna Lukic ◽  
Francesco de Gasperin ◽  
Marcus Brüggen
Keyword(s):  
Deep Learning ◽  
Galactic Nuclei ◽  
Correlated Noise ◽  
Artificial Data ◽  
Star Forming ◽  
Gaussian Fitting ◽  
Fitting Method ◽  
Novel Method ◽  
Square Kilometer Array ◽  
Total Integration

Finding and classifying astronomical sources is key in the scientific exploitation of radio surveys. Source-finding usually involves identifying the parts of an image belonging to an astronomical source, against some estimated background. This can be problematic in the radio regime, owing to the presence of correlated noise, which can interfere with the signal from the source. In the current work, we present ConvoSource, a novel method based on a deep learning technique, to identify the positions of radio sources, and compare the results to a Gaussian-fitting method. Since the deep learning approach allows the generation of more training images, it should perform well in the source-finding task. We test the source-finding methods on artificial data created for the data challenge of the Square Kilometer Array (SKA). We investigate sources that are divided into three classes: star forming galaxies (SFGs) and two classes of active galactic nuclei (AGN). The artificial data are given at two different frequencies (560 MHz and 1400 MHz), three total integration times (8 h, 100 h, 1000 h), and three signal-to-noise ratios (SNRs) of 1, 2, and 5. At lower SNRs, ConvoSource tends to outperform a Gaussian-fitting approach in the recovery of SFGs and all sources, although at the lowest SNR of one, the better performance is likely due to chance matches. The Gaussian-fitting method performs better in the recovery of the AGN-type sources at lower SNRs. At a higher SNR, ConvoSource performs better on average in the recovery of AGN sources, whereas the Gaussian-fitting method performs better in the recovery of SFGs and all sources. ConvoSource usually performs better at shorter total integration times and detects more true positives and misses fewer sources compared to the Gaussian-fitting method; however, it detects more false positives.

scholarly journals High-Order Polynomial Fitting Assistance for Fast Double-Peak Finding in Brillouin-Distributed Sensing

Sensors ◽  
2020 ◽  
Vol 21 (1) ◽  
pp. 187
Author(s):  
Marcelo A. Soto ◽  
Alin Jderu ◽  
Dorel Dorobantu ◽  
Marius Enachescu ◽  
Dominik Ziegler
Keyword(s):  
Optical Fibers ◽  
High Order ◽  
Polynomial Fitting ◽  
Gaussian Fitting ◽  
Peak Finding ◽  
Fitting Method ◽  
Fold Reduction ◽  
Order Polynomial ◽  
Fitting In ◽  
Distributed Brillouin Sensor

A high-order polynomial fitting method is proposed to accelerate the computation of double-Gaussian fitting in the retrieval of the Brillouin frequency shifts (BFS) in optical fibers showing two local Brillouin peaks. The method is experimentally validated in a distributed Brillouin sensor under different signal-to noise ratios and realistic spectral scenarios. Results verify that a sixth-order polynomial fitting can provide a reliable initial estimation of the dual local BFS values, which can be subsequently used as initial parameters of a nonlinear double-Gaussian fitting. The method demonstrates a 4.9-fold reduction in the number of iterations required by double-Gaussian fitting and a 3.4-fold improvement in processing time.

Deep Learning based Thermal Crack Detection on Structural Concrete Exposed to Elevated Temperature

2021 ◽  
pp. 136943322098663
Author(s):  
Diana Andrushia A ◽  
Anand N ◽  
Eva Lubloy ◽  
Prince Arulraj G
Keyword(s):  
Deep Learning ◽  
Crack Detection ◽  
Detection Method ◽  
Concrete Structures ◽  
Vital Role ◽  
Subjective Knowledge ◽  
Thermal Cracks ◽  
Reinforced Cement ◽  
Novel Method ◽  
Human Inspection

Health monitoring of concrete including, detecting defects such as cracking, spalling on fire affected concrete structures plays a vital role in the maintenance of reinforced cement concrete structures. However, this process mostly uses human inspection and relies on subjective knowledge of the inspectors. To overcome this limitation, a deep learning based automatic crack detection method is proposed. Deep learning is a vibrant strategy under computer vision field. The proposed method consists of U-Net architecture with an encoder and decoder framework. It performs pixel wise classification to detect the thermal cracks accurately. Binary Cross Entropy (BCA) based loss function is selected as the evaluation function. Trained U-Net is capable of detecting major thermal cracks and minor thermal cracks under various heating durations. The proposed, U-Net crack detection is a novel method which can be used to detect the thermal cracks developed on fire exposed concrete structures. The proposed method is compared with the other state-of-the-art methods and found to be accurate with 78.12% Intersection over Union (IoU).

scholarly journals Constraining photoionization models with a reprojected optical diagnostic diagram

2020 ◽  
Vol 499 (4) ◽  
pp. 5749-5764 ◽  
Author(s):  
Xihan Ji ◽  
Renbin Yan
Keyword(s):  
Star Formation ◽  
Active Galactic Nuclei ◽  
Galactic Nuclei ◽  
Theoretical Models ◽  
Line Ratio ◽  
Star Forming ◽  
Optical Diagnostic ◽  
Diagnostic Diagram ◽  
Upper Limits ◽  
Nitrogen Abundance

ABSTRACT Optical diagnostic diagrams are powerful tools to separate different ionizing sources in galaxies. However, the model-constraining power of the most widely used diagrams is very limited and challenging to visualize. In addition, there have always been classification inconsistencies between diagrams based on different line ratios, and ambiguities between regions purely ionized by active galactic nuclei (AGNs) and composite regions. We present a simple reprojection of the 3D line ratio space composed of [N ii]λ6583/H α, [S ii]λλ6716, 6731/H α, and [O iii]λ5007/H β, which reveals its model-constraining power and removes the ambiguity for the true composite objects. It highlights the discrepancy between many theoretical models and the data loci. With this reprojection, we can put strong constraints on the photoionization models and the secondary nitrogen abundance prescription. We find that a single nitrogen prescription cannot fit both the star-forming locus and AGN locus simultaneously, with the latter requiring higher N/O ratios. The true composite regions stand separately from both models. We can compute the fractional AGN contributions for the composite regions, and define demarcations with specific upper limits on contamination from AGN or star formation. When the discrepancy about nitrogen prescriptions gets resolved in the future, it would also be possible to make robust metallicity measurements for composite regions and AGNs.

Multiwavelength analysis of OH Megamaser galaxies: The case of IRAS11506-3851

2020 ◽  
Vol 15 (S359) ◽  
pp. 347-349
Author(s):  
Carpes P. Hekatelyne ◽  
Thaisa Storchi-Bergmann
Keyword(s):  
Active Galactic Nuclei ◽  
Hubble Space Telescope ◽  
Galactic Nuclei ◽  
Large Array ◽  
Very Large Array ◽  
Star Forming ◽  
Star Forming Regions ◽  
Field Unit ◽  
Mixed Transition ◽  
Integral Field

AbstractWe present Multi-Object Spectrograph (GMOS) Integral Field Unit (IFU), Hubble Space Telescope (HST) and Very Large Array (VLA) observations of the inner kpc of the OH Megamaser galaxy IRAS 11506-3851. In this work we discuss the kinematics and excitation of the gas as well as its radio emission. The HST images reveal an isolated spiral galaxy and the combination with the GMOS-IFU flux distributions allowed us to identify a partial ring of star-forming regions surrounding the nucleus with a radius of ≍500 pc. The emission-line ratios and excitation map reveal that the region inside the ring present mixed/transition excitation between those of Starbursts and Active Galactic Nuclei (AGN), while regions along the ring are excited by Starbursts. We suggest that we are probing a buried or fading AGN that could be both exciting the gas and originating an outflow.

scholarly journals A Novel Method of Billing System Using Deep Learning

2021 ◽  
Vol 1084 (1) ◽  
pp. 012035
Author(s):  
J Haritha ◽  
K Prakash ◽  
B Navina ◽  
S Saveetha
Keyword(s):  
Deep Learning ◽  
Billing System ◽  
Novel Method

scholarly journals X-ray properties of He ii λ 1640 emitting galaxies in VANDELS

2020 ◽  
Vol 496 (3) ◽  
pp. 3796-3807
Author(s):  
A Saxena ◽  
L Pentericci ◽  
D Schaerer ◽  
R Schneider ◽  
R Amorin ◽  
...  
Keyword(s):  
Formation Rate ◽  
Galactic Nuclei ◽  
X Ray ◽  
Star Forming ◽  
Set Constraints ◽  
Significant Difference ◽  
Low Metallicity ◽  
Matched Samples ◽  
X Ray Binaries

ABSTRACT We explore X-ray emission from a sample of 18 He ii λ1640 emitting star-forming galaxies at z ∼ 2.3–3.6 from the VANDELS survey in the Chandra Deep Field South, to set constraints on the role of X-ray sources in powering the He ii emission. We find that 4 He ii emitters have tentative detections with S/N ∼ 2 and have X-ray luminosities, LX = 1.5−4.9 × 1041 erg s−1. The stacked luminosity of all 18 He ii emitters is 2.6 × 1041 erg s−1, and that of a subset of 13 narrow He ii emitters (FHWM(He ii) < 1000 km s−1) is 3.1 × 1041 erg s−1. We also measure stacked LX for non-He ii emitters through bootstrapping of matched samples, and find LX = 2.5 × 1041 erg s−1, which is not significantly different from LX measured for He ii emitters. The LX per star formation rate for He ii emitters (log (LX/SFR) ∼ 40.0) and non-emitters (log (LX/SFR) ∼ 39.9) are also comparable and in line with the redshift evolution and metallicity dependence predicted by models. Due to the non-significant difference between the X-ray emission from galaxies with and without He ii, we conclude that X-ray binaries or weak or obscured active galactic nuclei are unlikely to be the dominant producers of He ii ionizing photons in VANDELS star-forming galaxies at z ∼ 3. Given the comparable physical properties of both He ii emitters and non-emitters reported previously, alternative He ii ionizing mechanisms such as localized low-metallicity stellar populations, Pop-III stars, etc. may need to be explored.

AMBIPOLAR DIFFUSION AND TURBULENT MAGNETIC FIELDS IN MOLECULAR CLOUDS

2011 ◽  
Vol 26 (04) ◽  
pp. 235-249 ◽  
Author(s):  
MARTIN HOUDE ◽  
TALAYEH HEZAREH ◽  
HUA-BAI LI ◽  
THOMAS G. PHILLIPS
Keyword(s):  
Magnetic Fields ◽  
Molecular Clouds ◽  
Ambipolar Diffusion ◽  
Star Forming ◽  
Star Forming Regions ◽  
Spin Interactions ◽  
Line Narrowing ◽  
Weakly Ionized ◽  
Novel Method

We review the introduction and development of a novel method for the characterization of magnetic fields in star-forming regions. The technique is based on the comparison of spectral line profiles from coexistent neutral and ion molecular species commonly detected in molecular clouds, sites of star formation. Unlike other methods used to study magnetic fields in the cold interstellar medium, this ion/neutral technique is not based on spin interactions with the field. Instead, it relies on and takes advantage of the strong cyclotron coupling between the ions and magnetic fields, thus exposing what is probably the clearest observational manifestation of magnetic fields in the cold, weakly ionized gas that characterizes the interior of molecular clouds. We will show how recent development and modeling of the ensuing ion line narrowing effect leads to a determination of the ambipolar diffusion scale involving the turbulent component of magnetic fields in star-forming regions, as well as the strength of the ordered component of the magnetic field.

A novel method based on machine vision system and deep learning to detect fraud in turmeric powder

2021 ◽  
pp. 104728
Author(s):  
Ahmad Jahanbakhshi ◽  
Yousef Abbaspour-Gilandeh ◽  
Kobra Heidarbeigi ◽  
Mohammad Momeny
Keyword(s):  
Deep Learning ◽  
Machine Vision ◽  
Vision System ◽  
Machine Vision System ◽  
Novel Method
Sign in / Sign up

Export Citation Format

Share Document