arXiv subjects

Shang Li

Publications and source records attributed to Shang Li.

At least 37 records · Page 2Linked to original sources

arXiv2022

From General to Specific: Online Updating for Blind Super-Resolution

Most deep learning-based super-resolution (SR) methods are not image-specific: 1) They are trained on samples synthesized by predefined degradations (e.g. bicubic downsampling), regardless of the domain gap between training and testing data. 2) During testing, they super-resolve all images by the same set of model weights, ignoring the degradation variety. As a result, most previous methods may suffer a performance drop when the degradations of test images are unknown and various (i.e. the case of blind SR). To address these issues, we propose an online SR (ONSR) method. It does not rely on predefined degradations and allows the model weights to be updated according to the degradation of the test image. Specifically, ONSR consists of two branches, namely internal branch (IB) and external branch (EB). IB could learn the specific degradation of the given test LR image, and EB could learn to super resolve images degraded by the learned degradation. In this way, ONSR could customize a specific model for each test image, and thus get more robust to various degradations. Extensive experiments on both synthesized and real-world images show that ONSR can generate more visually favorable SR results and achieve state-of-the-art performance in blind SR.

cs.CV↗

arXiv2021

Search for gamma-ray emission from the 12 nearby dwarf spheroidal galaxies with 12 years of Fermi-LAT data

Previously, we have shown in Li et al. (2018) that very weak $γ$-ray excesses ($\sim 2σ$) appear in some Milky Way dwarf spheroidal galaxies (dSphs, including candidates) and the combination analysis of 12 nearby dSphs yields a local significance of $>4σ$. In this work, we adopt a longer data set (i.e., the 12 years of Fermi-LAT data), the latest Fermi-LAT software as well as background models to update the searches of $γ$-ray emission from these sources. Very weak $γ$-ray excesses ($>$ 2 $σ$) are found in the directions of three dSphs, including Reticulum II, Bootes II and Willman 1. In the direction of Reticulum II, the peak TS value of the excess reaches $\sim$11. However, different from the previous analysis with the 9 years of Fermi-LAT data, now the location of the gamma-ray emission is significantly away from the center of Reticulum II because of the enhancement of "offset" $γ$-rays above 10 GeV since 2017. The detected weak excess is likely due to the contamination of an astrophysical $γ$-ray source with a TS value of $\sim 22$, irrelevant to the dark matter inside Reticulum II. The possible excesses in the directions of Bootes II and Willman 1 are weaker with lower peak TS values ($\sim 7$). If interpreted as annihilation of dark matter particles into $τ^+τ^-$, the dark mass mass of $m_χ\sim 14$ GeV and $\sim 80$ GeV are found for Bootes II and Willman 1, respectively. Much more data are needed to clarify whether these two potential signals are real and then reveal their origins.

astro-ph.HE↗

arXiv2021

The Highly Self-Absorbed Blazar, PKS\,1351$-$018

PKS\,1351$-$018 at a redshift of $z=3.71$ is one of the most luminous, steady synchrotron sources with a luminosity $> 10^{47}$\,erg~s$^{-1}$. The synchrotron luminosity does not seem to vary by more than $\sim 25\%$ over 35 years. In order to appreciate this remarkable behavior, if it were at $z=0.5$, it would have a flux density at 15 GHz in a range of $110 - 137$\,Jy over 11 yrs. In spite of this steady behavior, two strong $γ$-ray flares $\lesssim 10^{49}$\,erg~s$^{-1}$ were detected in 2011 and 2016. There is blazar-like behavior coexisting with the steady behavior. This study is aimed at elucidating the dual nature of this source. We find that the radio source is extremely compact with a bright core and a steep spectrum secondary component, 12\,mas away, that appears to be constant in position and flux density in six epochs from 1995 to 2018. We estimate that a jet with a time averaged power of $(5.2 \pm 3.2) \times 10^{45}$\,erg~s$^{-1}$ terminates in this lobe that is advancing $\gtrsim 0.9 c$ at a deprojected distance of $1-3$\,kpc from the central engine. This is the rare case of a young ($\sim 6000$\,yr), very powerful radio source that is viewed a few degrees from the jet axis. We find evidence of a high velocity (4000\,km~s$^{-1}$), high ionization wind emanating form a luminous quasar. The young radio jet appears to experience modest bending as it navigates through the intense quasar environment.

astro-ph.HE↗

arXiv2021

End-to-end Alternating Optimization for Blind Super Resolution

Previous methods decompose the blind super-resolution (SR) problem into two sequential steps: \textit{i}) estimating the blur kernel from given low-resolution (LR) image and \textit{ii}) restoring the SR image based on the estimated kernel. This two-step solution involves two independently trained models, which may not be well compatible with each other. A small estimation error of the first step could cause a severe performance drop of the second one. While on the other hand, the first step can only utilize limited information from the LR image, which makes it difficult to predict a highly accurate blur kernel. Towards these issues, instead of considering these two steps separately, we adopt an alternating optimization algorithm, which can estimate the blur kernel and restore the SR image in a single model. Specifically, we design two convolutional neural modules, namely \textit{Restorer} and \textit{Estimator}. \textit{Restorer} restores the SR image based on the predicted kernel, and \textit{Estimator} estimates the blur kernel with the help of the restored SR image. We alternate these two modules repeatedly and unfold this process to form an end-to-end trainable network. In this way, \textit{Estimator} utilizes information from both LR and SR images, which makes the estimation of the blur kernel easier. More importantly, \textit{Restorer} is trained with the kernel estimated by \textit{Estimator}, instead of the ground-truth kernel, thus \textit{Restorer} could be more tolerant to the estimation error of \textit{Estimator}. Extensive experiments on synthetic datasets and real-world images show that our model can largely outperform state-of-the-art methods and produce more visually favorable results at a much higher speed. The source code is available at \url{https://github.com/greatlog/DAN.git}.

cs.CV↗

arXiv2021

Determination of local defect density in diamond by double electron-electron resonance

Magnetic impurities in diamond influence the relaxation properties and thus limit the sensitivity of magnetic, electric, strain, and temperature sensors based on nitrogen-vacancy color centers. Diamond samples may exhibit significant spatial variations in the impurity concentrations hindering the quantitative analysis of relaxation pathways. Here, we present a local measurement technique which can be used to determine the concentration of various species of defects by utilizing double electron-electron resonance. This method will help to improve the understanding of the physics underlying spin relaxation and guide the development of diamond samples, as well as offering protocols for optimized sensing.

cond-mat.mes-hall↗

arXiv2020

Unfolding the Alternating Optimization for Blind Super Resolution

Previous methods decompose blind super resolution (SR) problem into two sequential steps: \textit{i}) estimating blur kernel from given low-resolution (LR) image and \textit{ii}) restoring SR image based on estimated kernel. This two-step solution involves two independently trained models, which may not be well compatible with each other. Small estimation error of the first step could cause severe performance drop of the second one. While on the other hand, the first step can only utilize limited information from LR image, which makes it difficult to predict highly accurate blur kernel. Towards these issues, instead of considering these two steps separately, we adopt an alternating optimization algorithm, which can estimate blur kernel and restore SR image in a single model. Specifically, we design two convolutional neural modules, namely \textit{Restorer} and \textit{Estimator}. \textit{Restorer} restores SR image based on predicted kernel, and \textit{Estimator} estimates blur kernel with the help of restored SR image. We alternate these two modules repeatedly and unfold this process to form an end-to-end trainable network. In this way, \textit{Estimator} utilizes information from both LR and SR images, which makes the estimation of blur kernel easier. More importantly, \textit{Restorer} is trained with the kernel estimated by \textit{Estimator}, instead of ground-truth kernel, thus \textit{Restorer} could be more tolerant to the estimation error of \textit{Estimator}. Extensive experiments on synthetic datasets and real-world images show that our model can largely outperform state-of-the-art methods and produce more visually favorable results at much higher speed. The source code is available at https://github.com/greatlog/DAN.git.

cs.CV↗

arXiv2020

Awakening of two $γ$-ray high redshift flat-spectrum radio quasars in the southern hemisphere

High-redshift blazars are valuable tools to study the early Universe. So far only a handful of $γ$-ray blazars have been found at redshifts above 3. $γ$-ray signals are detected in the direction of PMN J2219-2719 ($z=3.63$) and PMN J2321-0827 ($z=3.16$) by analyzing the 10-year $Fermi$-LAT Pass 8 data. PMN J2219-2719 is not distinguished from the background in the global analysis. During the 5-month epoch, the TS value is 47.8 and the flux is more than 10 times of the 10-year averaged flux. In addition, the angular distance between the $γ$-ray position and the radio position of PMN J2219-2719 is only ${0.04}^{\circ}$. Moreover, the $γ$-ray and infrared light curves of long time scale are very similar, which support the association between the $γ$-ray source and PMN J2219-2719. The global analysis of PMN J2321-0827 suggest a new $γ$-ray source, during the flare phase, the TS value is 61.4 and the $γ$-ray flux increased significantly. The association probability suggests that PMN J2321-0827 may be the counterpart of the new $γ$-ray source. In the future, the number of high-redshift $γ$-ray sources will increase by combining $Fermi$-LAT and the upcoming Large Synoptic Survey Telescope.

astro-ph.HE↗

arXiv2020

Detections of simultaneous brightening of gamma-ray and optical emissions of a distant blazar GB 1508+5714 at redshift 4.3

GB 1508+5714 is a high-redshift blazar ($z$ = 4.3) and a spectrally soft $γ$-ray source has been detected in its direction. By analyzing 11.4-yr {\it Fermi}-LAT data, significant long-term variability of the $γ$-ray source is confirmed. More importantly, a $γ$-ray emission flare appeared in an epoch of several tens of days in year 2018, when the flux is about four times of the value from the global fit. Meanwhile, optical flares displayed in both $r$- and $i$-bands from the {\it Zwicky} Transient Facility light curves. Detections of the simultaneous $γ$-ray and optical brightening provide a decisive evidence to pin down the association between the $γ$-ray source and GB 1508+5714, which makes it the {\it first} identified $γ$-ray blazar beyond redshift 4. A broadband spectral energy distribution in the high flux state is constructed and the origin of the multiwavelength brightening is also briefly discussed. Upcoming wide-deep-fast optical telescopes together with the $γ$-ray sky surveyors will shed lights on the role that the AGN jets play in the early cosmic time.

astro-ph.HE↗

arXiv2019

Search for line signal candidates in the Fermi-LAT data

In order to search for the line-like signals in the Fermi-LAT data, we have analyzed totally 49152 regions of interest (ROIs) that cover the whole sky. No ROI displays a line signal with test statistic (TS) value above 25, while for 50 ROIs weak line-like excesses with ${\rm TS}>16$ are presented. The intrinsic significances of these potential signals are further reduced by the large trial factor introduced in such kind of analysis. For the largest TS value of 24.3 derived in our analysis, the corresponding global significance is only $0.54σ$. We thus do not find any significant line-like signal and set up constraints on the cross section of dark matter annihilating to gamma-ray lines, $\left<σ{v}\right>_{γγ}$.

astro-ph.HE↗

arXiv2018

$Fermi$-LAT detection of a transient $γ$-ray source in the direction of a distant blazar B3 1428+422 at $z =4.72$

We report the detection of a transient $γ$-ray source in the direction of B3 1428+422 ($z=4.72$) by analyzing the 110-month {\it Fermi}-LAT Pass 8 data. The new transient $γ-$ray source is far away from the Galactic plane and has a rather soft spectrum, in agreement with being a high redshift blazar. We suggest that the newly discovered transient is the $γ$-ray counterpart of B3 1428+422, which could be the {\it most distant} GeV source detected so far. The detection of a group of such distant $γ-$ray blazars will be helpful to reconstruct the evolution of the luminosity function and to study the extragalactic background light at such high redshifts.

astro-ph.HE↗

arXiv2018

Active galactic nuclei with GeV activities and the PeV neutrino source candidate TXS 0506+056

On 2017 September 22 the IceCube neutrino observatory detected a track-like, very-high-energy event (IceCube-170922A) that is spatially associated with TXS 0506+056, a quasar at a redshift of $z=0.3365$. This source is characterized by the increased acitivies in a very wide energy range (from radio to TeV) during these days. To investigate the possible connection of the PeV neutrino emission with the GeV activity of blazars, in this work we select 116 bright sources and analyze their lightcurves and spectra. We focus on the sources displaying GeV activities. Among these blazars, TXS 0506+056 seems to be typical in many aspects but is distinguished by the very strong GeV activties. We suggest to search for neutrino outburst in the historical data of IceCube, as recently done for TXS 0506+056, from the directions of these more energetic and harder blazars with strong GeV activities.

astro-ph.HE↗

arXiv2018

Search for gamma-ray emission from the nearby dwarf spheroidal galaxies with 9 years of Fermi-LAT data

In this work, we search for $γ$-ray emission from the directions of some nearby Milky Way dwarf spheroidal galaxies (dSphs) and candidates with the publicly-available Pass 8 data of Fermi-LAT. Our sample includes 12 sources with the distances $<50$ kpc. Very weak $γ$-ray excesses ($\sim 2σ$) are found in some dSphs/candidates, consistent with that reported in the previous literature. Intriguingly, the peak test statistic (TS) value of the weak emission from the direction of Reticulum II rises continually. If interpreted as dark matter (DM) annihilation, the peak TS value is 13.5 for the annihilation channel of $χχ\rightarrow τ^{+}τ^{-}$ and the DM mass of $m_χ\sim 16$ GeV. The combination of all these nearby sources yields a more significant (with local significance $> 4σ$) $γ$-ray signal.

astro-ph.HE↗

arXiv2018

New Bounds on Axion-Like Particles From the Fermi Large Area Telescope observation of PKS $2155-304$

The axion-like particle (ALP)-photon mixing in the magnetic field around $γ$-ray sources or along the line-of-sight could induce oscillation between photons and ALPs, which then causes irregularities in the $γ$-ray spectra. In this work we try to search for such spectral irregularities in the spectrum of PKS $2155-304$ using 8.6 years of the Fermi Large Area Telescope (Fermi-LAT) data. No significant evidence for the presence of ALP-photon oscillation is obtained, and the parameter space of ALPs is constrained. The exclusion region sensitively depends on the poorly known magnetic field of host galaxy cluster of PKS $2155-304$. If the magnetic field is as high as $\sim 10~{\rm μG}$, the "hole"-like parameter region allowed in Ref.~\cite{Fermi} can be ruled out.

hep-ph↗

arXiv2018

Is the gamma-ray emission around the Coma galaxy cluster a dark matter origin?

Recently, gamma-ray emission in the direction of Coma, with a TS value of $\sim 40$, has been reported. In this work we will discuss the possibility of such a residual emission coming from dark matter annihilation. Our results show that the gamma-ray emission within the Coma region is also spatially correlated to the mass distribution derived from weak gravitational lensing measurements very well. However the dark matter models are not supported by the spectral analysis results and constraints by observations of other targets. Thus we derive the upper limits of the dark matter annihilation cross section according to the observation of the Coma region.

astro-ph.HE↗

arXiv2017

Fast $γ$-ray variability in blazars beyond redsift 3

High-redshift blazars are one of the most powerful monsters in the universe and $γ$-ray variability carries crucial information of their relativistic jets. In this work we present results of the first systematical temporal analysis of {\it Fermi}-LAT data of all known seven $γ$-ray blazars beyond redshift 3. Significant long-term $γ$-ray variability is found from five sources in monthly $γ$-ray light curves, in which three of them are reported for the first time. Furthermore, intraday $γ$-ray variations are detected from NVSS J053954$-$283956 and NVSS J080518$+$614423. Doubling variability timescale of the former source is limited as short as $\lesssim$ 1 hour (at the source frame). Together with variability amplitude over one order of magnitude, NVSS J053954$-$283956 is the most distant $γ$-ray flaring blazar so far. Meanwhile, intraday optical variability of NVSS J163547$+$362930 is found based on archival PTF/iPTF light curve. Benefited from multiwavelength activity of these sources, constraints of their Doppler factors as well as locations of $γ$-ray radiation region and indications for the SDSS high redshift jetted active galactic nuclei deficit are discussed.

astro-ph.HE↗

arXiv2017

3FGL J1924.8-1034 : A Spatially-Extended Stable Unidentified GeV Source?

Milky Way-like galaxies are predicted to host a very large number of dark matter subhalos. Some massive and nearby subhalos could generate detectable gamma-rays, appearing as unidentified, spatially-extended and stable gamma-ray sources. We search for such sources in the third Fermi Large Area Telescope source List (3FGL) and report the identification of a new candidate, 3FGL J1924.8-1034. With the Fermi-LAT Pass 8 data, we find that 3FGL J1924.8-1034 is spatially-extended at a high confidence level of $5.4σ$, with a best-fit extension radius of $\sim0.15^{\circ}$. No significant variability has been found and its gamma-ray spectrum is well fitted by the dark matter annihilation into $b\bar{b}$ with a mass of $\sim 43$ GeV. All these facts make 3FGL J1924.8-1034 a possible dark matter subhalo candidate. However, due to the limited angular resolution, the possibility of that the spatial extension of 3FGL J1924.8-1034 is caused by the contamination from the other un-resolved point source can not be ruled out.

astro-ph.HE↗

arXiv2016

Testing the dark matter subhalo hypothesis of the gamma-ray source 3FGL J2212.5+0703

N-body simulations predict that galaxies at the Milky Way scale host a large number of dark matter (DM) subhalos. Some of these subhalos, if they are massive enough or close enough to the Earth, might be detectable in $γ$ rays due to the DM annihilation. 3FGL J2212.5+0703, an unidentified gamma-ray source, has been suggested to be the counterpart candidate of a DM subhalo by Bertoni et al. (2015, 2016). In this work we analyze the Fermi-LAT Pass 8 data of 3FGL J2212.5+0703 to independently test the DM subhalo hypothesis of this source. In order to suppress the possible contamination from two nearby very-bright blazars, we just take into account the front-converting gamma-rays which have better angular resolutions than that of the back-converting photons. In addition to the spatial distribution analysis, we have extended the spectrum analysis down to the energies of $\sim 100$ MeV, and thoroughly examined the variability of the emission during the past 8 years. We confirm that 3FGL J2212.5+0703 is a steady and spatially-extended gamma-ray emitter at a high confidence level. The spectrum is well consistent with that expected from DM annihilation into $b\bar{b}$. The introduction of a phenomenological LogParabola spectrum just improves the fit slightly. All these results suggest that 3FGL J2212.5+0703 could be indicative of a DM subhalo.

astro-ph.HE↗

arXiv2016

Order-2 Asymptotic Optimality of the Fully Distributed Sequential Hypothesis Test

This work analyzes the asymptotic performances of fully distributed sequential hypothesis testing procedures as the type-I and type-II error rates approach zero, in the context of a sensor network without a fusion center. In particular, the sensor network is defined by an undirected graph, where each sensor can observe samples over time, access the information from the adjacent sensors, and perform the sequential test based on its own decision statistic. Different from most literature, the sampling process and the information exchange process in our framework take place simultaneously (or, at least in comparable time-scales), thus cannot be decoupled from one another. Two message-passing schemes are considered, based on which the distributed sequential probability ratio test (DSPRT) is carried out respectively. The first scheme features the dissemination of the raw samples. Although the sample propagation based DSPRT is shown to yield the asymptotically optimal performance at each sensor, it incurs excessive inter-sensor communication overhead due to the exchange of raw samples with index information. The second scheme adopts the consensus algorithm, where the local decision statistic is exchanged between sensors instead of the raw samples, thus significantly lowering the communication requirement compared to the first scheme. In particular, the decision statistic for DSPRT at each sensor is updated by the weighted average of the decision statistics in the neighbourhood at every message-passing step. We show that, under certain regularity conditions, the consensus algorithm based DSPRT also yields the order-2 asymptotically optimal performance at all sensors.

stat.AP↗