Search

Magnetar is proposed as one of the possible central engines for a gamma-ray burst (GRB). Recent studies show that if a magnetar has a rotational axis misaligned from the magnetic one, a periodic lightcurve pattern is expected with a period of seconds to minutes. Inspired by this unique feature, in this paper, we search for the quasi-periodic oscillation (QPO) signals in the {\it Swift} observations of GRBs. Using the Lomb-Scargle periodogram and the weighted wavelet Z-transform algorithms, we find that the {\it Swift}-BAT data of GRB 210514A has a QPO signal with a period $\sim 11\,{\rm s}$. The estimated confidence level of the signal is over $3 \sigma$. The global lightcurve of this GRB exhibits a double-plateau structure with a sharp decay segment between plateaus. The lightcurve feature resembles those of GRBs that were reported to have internal plateaus. We explain the observations of GRB 210514A with a supra-massive magnetar (SMM) model, where the QPO signal in the first plateau is produced via the dipole radiation of the SMM experiencing a precession motion, the sharp decay is due to the collapse of the SMM into a black hole (BH), and the second plateau could be produced via the fall-back accretion of the newborn BH. We fit the precession model to the observations using the Bayesian statistic and the best-fit magnetar parameters are discussed. Alternative models concerning a BH central engine may also provide reasonable explanations for this burst, only in this case the QPO signal could merely be a coincidence., Comment: 13 pages, 4 figures, accepted for publication in ApJ

A good feature representation is the key to image classification. In practice, image classifiers may be applied in scenarios different from what they have been trained on. This so-called domain shift leads to a significant performance drop in image classification. Unsupervised domain adaptation (UDA) reduces the domain shift by transferring the knowledge learned from a labeled source domain to an unlabeled target domain. We perform feature disentanglement for UDA by distilling category-relevant features and excluding category-irrelevant features from the global feature maps. This disentanglement prevents the network from overfitting to category-irrelevant information and makes it focus on information useful for classification. This reduces the difficulty of domain alignment and improves the classification accuracy on the target domain. We propose a coarse-to-fine domain adaptation method called Domain Adaptation via Feature Disentanglement~(DAFD), which has two components: (1)the Category-Relevant Feature Selection (CRFS) module, which disentangles the category-relevant features from the category-irrelevant features, and (2)the Dynamic Local Maximum Mean Discrepancy (DLMMD) module, which achieves fine-grained alignment by reducing the discrepancy within the category-relevant features from different domains. Combined with the CRFS, the DLMMD module can align the category-relevant features properly. We conduct comprehensive experiment on four standard datasets. Our results clearly demonstrate the robustness and effectiveness of our approach in domain adaptive image classification tasks and its competitiveness to the state of the art., Comment: Update the experimental results

The observed early X-ray plateau in the afterglow lightcurves of some gamma-ray bursts (GRBs) is attributed to the dipole radiations (DRs) of a newborn magnetar. A quasi-periodic oscillation (QPO) signal in the plateau would be strong evidence of the magnetar precession motion. By making a time-frequency domain analysis for the X-ray afterglow lightcurve of GRB 180620A, we find a QPO signal of $\sim650$ seconds in its early X-ray plateau. We fit the lightcurve with a magnetar precession model by adopting the Markov chain Monte Carlo algorithm. The observed lightcurve and the QPO signal are well represented with our model. The derived magnetic field strength of the magnetar is $B_{\rm p}= (1.02^{+0.59}_{-0.61})\times10^{15}$~G. It rapidly spins down with angular velocity evolving as $\Omega_{s} \propto(1+t/\tau_{\rm sd})^{-0.96}$, where $\tau_{\rm sd}=9430$~s. Its precession velocity evolution is even faster than $\Omega_s$, i.e. $\Omega_{ p}\propto (1+t/\tau_{p})^{-2.18\pm0.11}$, where $\tau_{p}=2239\pm206$~s. The inferred braking index is $n=2.04$. We argue that the extra energy loss via the magnetospheric processes results in its rapid spin-down, a low braking index of the magnetar, and the strong precession motion., Comment: 5 pages, 3 figures, accepted for publication in MNRAS Letters

Household garbage images are usually faced with complex backgrounds, variable illuminations, diverse angles, and changeable shapes, which bring a great difficulty in garbage image classification. Due to the ability to discover problem-specific features, deep learning and especially convolutional neural networks (CNNs) have been successfully and widely used for image representation learning. However, available and stable household garbage datasets are insufficient, which seriously limits the development of research and application. Besides, the state of the art in the field of garbage image classification is not entirely clear. To solve this problem, in this study, we built a new open benchmark dataset for household garbage image classification by simulating different lightings, backgrounds, angles, and shapes. This dataset is named 30 Classes of Household Garbage Images (HGI-30), which contains 18,000 images of 30 household garbage classes. The publicly available HGI-30 dataset allows researchers to develop accurate and robust methods for household garbage recognition. We also conducted experiments and performance analysis of the state-of-the-art deep CNN methods on HGI-30, which serves as baseline results on this benchmark.

Object recognition from images means to automatically find object(s) of interest and to return their category and location information. Benefiting from research on deep learning, like convolutional neural networks~(CNNs) and generative adversarial networks, the performance in this field has been improved significantly, especially when training and test data are drawn from similar distributions. However, mismatching distributions, i.e., domain shifts, lead to a significant performance drop. In this paper, we build domain-invariant detectors by learning domain classifiers via adversarial training. Based on the previous works that align image and instance level features, we mitigate the domain shift further by introducing a domain adaptation component at the region level within Faster \mbox{R-CNN}. We embed a domain classification network in the region proposal network~(RPN) using adversarial learning. The RPN can now generate accurate region proposals in different domains by effectively aligning the features between them. To mitigate the unstable convergence during the adversarial learning, we introduce a balanced domain classifier as well as a network learning rate adjustment strategy. We conduct comprehensive experiments using four standard datasets. The results demonstrate the effectiveness and robustness of our object detection approach in domain shift scenarios., Comment: fixed some issues

Assuming that the shallow-decaying phase in the early X-ray lightcurves of gamma-ray bursts (GRBs) is attributed to the dipole radiations (DRs) of a newborn magnetar, we present a comparative analysis for the magnetars born in death of massive stars and merger of compact binaries with long and short GRB (lGRB and sGRB) data observed with the {\em Swift} mission. We show that the typical braking index ($n$) of the magnetars is $\sim 3$ in the sGRB sample, and it is $\sim 4$ for the magnetars in the lGRB sample. Selecting a sub-sample of the magnetars whose spin-down is dominated by DRs ($n\lesssim 3$) and adopting a universal radiation efficiency of $0.3$, we find that the typical magnetic field strength ($B_p$) is $10^{16}$ G {\em vs.} $10^{15}$ G and the typical initial period ($P_0$) is $\sim 20$ ms {\em vs.} $2$ ms for the magnetars in the sGRBs {\em vs.} lGRBs. They follow the same relation between $P_0$ and the isotropic GRB energy as $ P_0\propto E_{\rm jet}^{-0.4}$. We also extend our comparison analysis to superluminous supernovae (SLSNe) and stable pulsars. Our results show that a magnetar born in merger of compact stars tends to have a stronger $B_p$ and a longer $P_0$ by about one order of magnitude than that born in collapse of massive stars. Its spin-down is dominated by the magnetic DRs as old pulsars, being due to its strong magnetic field strength, whereas the early spin-down of magnetars born in massive star collapse is governed by both the DRs and gravitational wave (GW) emission. A magnetar with a faster rotation speed should power a more energetic jet, being independent of its formation approach., Comment: 10 pages, 2 tables, 7 figures. Published in MNRAS

The unusual multiwavelength lightcurves of GRB 101225A are revisited by assuming that it is from an off-axis GRB powered by a newborn magnetar. We show that its optical afterglow lightcurve is fitted with the forward shock model by parameterizing its jet structure as a Gaussian function with a half opening angle of the jet core as $1.67^{\rm o}$. The derived initial Lorentz factor ($\Gamma_0$) is 120, and the viewing angle to the jet axis is $\theta_v=3.7^{\rm o}$. Tentative QPO signatures of $P=488$ seconds and $P=250\sim 300$ seconds are found with a confidence level of 90\% by analysing its X-ray flares observed in the time interval of $[4900,\ 7500]$ seconds. Its global gamma-ray/X-ray lightcurve and the QPO signatures are represented with the magnetar dipole radiation (DR) model by considering the magnetar precession motion, assuming that the magnetar spindown is dominated by the GW emission. The bulk Lorentz factor of the DR ejecta is limited to 8, being much lower than $\Gamma_0$. Comparing GRB 101225A with the extremely off-axis GRB 170817A, we suspect that the nature of the two-component jet in GRB 170817A is a combination of a co-axial GRB jet and a DR ejecta. GRB 101225A would be among the brightest ones of the CDF-S XT2 like X-ray transient population driven by newborn magnetars. Discussion on detectability of its gravitational wave emission is also presented., Comment: 8 pages,4 figures. Accepted for publication in ApJL on 10/12/2021

The X-ray flares have usually been ascribed to long-lasting activities of the central engine of gamma-ray bursts (GRBs), e.g., fallback accretion. The GRB X-ray plateaus, however, favor a millisecond magnetar central engine. The fallback accretion can be significantly suppressed due to the propeller effect of a magnetar. Therefore, if the propeller regime cannot resist the mass flow onto the surface of the magnetar efficiently, the X-ray flares raise upon the magnetar plateau would be hinted. In this work, such peculiar cases are connected to the accretion process of a magnetar, and an implication for magnetar-disc structure is given. We investigate the repeating accretion process with multi-flare GRB 050730, and give a discussion for the accreting induced variation of the magnetic field in GRB 111209A. Two or more flares exhibit in the GRB 050730, GRB 060607A, and GRB 140304A; by adopting magnetar mass $M=1.4~ M_\odot$ and radius $R=12~\rm km$, the average mass flow rates of the corresponding surrounding disk are $3.53\times 10^{-4}~M_\odot~\rm s^{-1}$, $4.23\times 10^{-4}~M_\odot~\rm s^{-1}$, and $4.33\times 10^{-4}~M_\odot~\rm s^{-1}$, and the corresponding average sizes of the magnetosphere are $5.01~\rm \times10^{6} cm$, $6.45~\rm \times10^{6} cm$, and $1.09~\rm \times10^{7} cm$, respectively. A statistic analysis that contains 8 GRBs within 12 flares shows that the total mass loading in single flare is $\sim 2\times 10^{-5}~M_{\odot}$. In the lost mass of a disk, there are about 0.1% used to feed a collimated jet., Comment: 15 pages, 4 figures, 2 tables, to be published in RAA

Gamma-ray burst (GRB) 150910A was detected by {\it Swift}/BAT, and then rapidly observed by {\it Swift}/XRT, {\it Swift}/UVOT, and ground-based telescopes. We report Lick Observatory spectroscopic and photometric observations of GRB~150910A, and we investigate the physical origins of both the optical and X-ray afterglows, incorporating data obtained with BAT and XRT. The light curves show that the jet emission episode lasts $\sim 360$~s with a sharp pulse from BAT to XRT (Episode I). In Episode II, the optical emission has a smooth onset bump followed by a normal decay ($\alpha_{\rm R,2} \approx -1.36$), as predicted in the standard external shock model, while the X-ray emission exhibits a plateau ($\alpha_{\rm X,1} \approx -0.36$) followed by a steep decay ($\alpha_{\rm X,2} \approx -2.12$). The light curves show obvious chromatic behavior with an excess in the X-ray flux. Our results suggest that GRB 150910A is an unusual GRB driven by a newly-born magnetar with its extremely energetic magnetic dipole (MD) wind in Episode II, which overwhelmingly dominates the observed early X-ray plateau. The radiative efficiency of the jet prompt emission is $\eta_{\gamma} \approx 11\%$. The MD wind emission was detected in both the BAT and XRT bands, making it the brightest among the current sample of MD winds seen by XRT. We infer the initial spin period ($P_0$) and the surface polar cap magnetic field strength ($B_p$) of the magnetar as $1.02 \times 10^{15}~{\rm G} \leq B_{p} \leq 1.80 \times 10^{15}~{\rm G}$ and 1~ms $\leq P_{0}v\leq 1.77$~ms, and the radiative efficiency of the wind is $\eta_w \geq 32\%$., Comment: Accepted for publication in ApJ; 26 pages, 8 figures, 3 tables

Early shallow-decaying X-ray afterglows of gamma-ray bursts (GRBs) may be attributed to the dipole radiations of newly-born magnetars. Assuming that the GRB jets powered by magnetars are quasi-universal, we find that the jet structure can be parameterized as a uniform jet with a luminosity of $\log L_{\rm j}/{\rm erg\ s^{-1}}=52.68^{+0.76}_{-0.33}$ (1$\sigma$) and an opening angle $\theta_{\rm j}=2.10_{-1.28}^{+1.90}$ (50\% confidence level), surrounding by a power-law decay component with an index of ${-4.00^{+0.27}_{-0.37}}$ (1$\sigma$). The inferred local GRB rate is $\rho=9.6$ Gpc$^{-3}$ yr$^{-1}$ by including both the typical GRBs and LL-GRBs as the same population. The typical viewing angle is $3.3^{o}$, and may be $20^{o}\sim30^{o}$ for LL-GRBs. The X-ray luminosity function of the dipole radiation wind can be empirically described by a broken power-law function with indices $\beta_1=0.78^{+0.16}_{-0.15}$ and $\beta_2>1.6$ broken at $\log L_{b, w}/{\rm erg\ s^{-1}}=48.51^{+0.53}_{-0.65}$. In case of that the wind outflow is collimated and co-axial with the GRB jet, we find that the wind structure is similar to the GRB jet, i.e., $\log L_{\rm c, w}/{\rm erg\ s^{-1}}=48.38^{+0.30}_{-0.48}$, $\theta_{\rm c, w}={2.65^{o}}_{-1.19^{o}}^{+0.1.73^{o}}$, and $k_{\rm w}=4.57^{+1.21}_{-0.75}$. The observed correlation between the prompt gamma-ray luminosity and X-ray luminosity of the wind may be resulted from the viewing angle effect in such a jet-wind system. Discussion on survey with the X-ray instruments on board the {\em Einstein\ Probe} mission in the soft X-ray band for the jet and wind emission is also presented., Comment: 22 pages, 10 figures, and 1 table. Accepted for publication in ApJ

Recently, with the availability of cost-effective depth cameras coupled with real-time skeleton estimation, the interest in skeleton-based human action recognition is renewed. Most of the existing skeletal representation approaches use either the joint location or the dynamics model. Differing from the previous studies, we propose a new method called Denoising Autoencoder with Temporal and Categorical Constraints (DAE_CTC)} to study the skeletal representation in a view of skeleton reconstruction. Based on the concept of learning under privileged information, we integrate action categories and temporal coordinates into a stacked denoising autoencoder in the training phase, to preserve category and temporal feature, while learning the hidden representation from a skeleton. Thus, we are able to improve the discriminative validity of the hidden representation. In order to mitigate the variation resulting from temporary misalignment, a new method of temporal registration, called Locally-Warped Sequence Registration (LWSR), is proposed for registering the sequences of inter- and intra-class actions. We finally represent the sequences using a Fourier Temporal Pyramid (FTP) representation and perform classification using a combination of LWSR registration, FTP representation, and a linear Support Vector Machine (SVM). The experimental results on three action data sets, namely MSR-Action3D, UTKinect-Action, and Florence3D-Action, show that our proposal performs better than many existing methods and comparably to the state of the art.

Gamma-ray burst (GRB) central engines and jet production mechanisms are still open questions. Assuming that the shallow decay segments of canonical X-ray afterglow lightcurves of {\em Swift} GRBs are attributed to the magnetic dipole (MD) radiations of newly-born magnetars, we derive the parameters of the magnetars and explore their possible relations to jet and MD wind emission. We show that the magnetar initial spin period ($P_0$) are tightly correlated with the jet energy ($E_{\rm jet}$), which is almost proportional to the wind energy ($E_{\rm wind}$). Our least square fits yield $P_0\propto E^{-0.36\pm 0.03}_{\rm jet}$ and $E_{\rm wind}\propto E^{0.91\pm 0.07}_{\rm jet}$. These relations may imply that a magnetar with faster rotating speed can power a more energetic GRB, and energy partition between the jet and wind may be quasi-universal. Although the $P_0-E_{\rm jet}$ relation is driven by a few sub-energetic GRBs in our sample, our Monte Carlo simulation analysis shows that sample selection biases from instrumental flux limits and contaminations of the bright jet afterglows cannot make this correlation. Within this jet-wind paradigm, we propose that GRB 101225A-like ultra-long GRBs, whose prompt gamma-ray/X-ray lightcurves are featured as a long-lasting plateau with a sharp drop, may be the orphan MD wind emission being due to misalignment of their jet axis to the light of sight. Brief discussion on the orphan MD wind emission and its association with the gravitational wave radiation of newly-born magnetars is presented., Comment: 19 pages, 2 tables, 12 figures, accepted for publication in ApJ

A newly-born magnetar is thought to be central engine of some long gamma-ray bursts (GRBs). We investigate the evolution of the electromagnetic (EM) emission from the magnetic dipole (MD) radiation wind injected by spin-down of a newly-born magnetar via both quadrupole gravitational-wave (GW) and MD radiations. We show that the EM luminosity evolves as $L_{\rm em}\propto (1+t/\tau_c)^{\alpha}$, and $\alpha$ is $-1$ and $-2$ in the GW and MD radiation dominated scenarios, respectively. Transition from the GW to MD radiation dominated epoch may show up as a smooth break with slope changing from $-1$ to $-2$. If the magnetar collapses to a black hole before $\tau_c$, the MD radiation should be shut down, then the EM light curve should be a plateau followed by a sharp drop. The expected generic light curve in this paradigm is consistent with the canonical X-ray light curve of {\em Swift} long GRBs. The X-ray emission of several long GRBs are identified and interpreted as magnetar spin-down via GW or MD, as well as constrain the physical parameters of magnetar. The combination of MD emission and GRB afterglows may make the diversity of the observed X-ray light curves. This may interpret the observed chromatic behaviors of the X-ray and optical afterglow light curves and the extremely low detection rate of a jet-like break in the X-ray afterglow light curves of long GRBs., Comment: 7 pages, 2 figures, Accepted for publication in MNRAS