Search

We perform a statistical inference of the astrophysical population of binary black hole (BBH) mergers observed during the first two observing runs of Advanced LIGO and Advanced Virgo, including events reported in the GWTC-1 and IAS catalogs. We derive a novel formalism to fully and consistently account for events of arbitrary significance. We carry out a software injection campaign to obtain a set of mock astrophysical events subject to our selection effects, and use the search background to compute the astrophysical probabilities $p_{\rm astro}$ of candidate events for several phenomenological models of the BBH population. We emphasize that the values of $p_{\rm astro}$ depend on both the astrophysical and background models. Finally, we combine the information from individual events to infer the rate, spin, mass, mass-ratio and redshift distributions of the mergers. The existing population does not discriminate between random spins with a spread in the effective spin parameter, and a small but nonzero fraction of events from tidally-torqued stellar progenitors. The mass distribution is consistent with one having a cutoff at $m_{\rm max} = 41^{+10}_{-5}\,\rm M_\odot$, while the mass ratio favors equal masses; the mean mass ratio $\bar q> 0.67$. The rate shows no significant evolution with redshift. We show that the merger rate restricted to BBHs with a primary mass between 20 and $30\, \rm M_\odot$, and a mass ratio $q > 0.5$, and at $z \sim 0.2$, is 1.5 to $5.3\,{\rm Gpc^{-3} yr^{-1}}$ (90\% c.l.); these bounds are model independent and a factor of $\sim 3$ tighter than that on the local rate of all BBH mergers, and hence are a robust constraint on all progenitor models. Including the events in our catalog increases the Fisher information about the BBH population by $\sim 47\%$, and tightens the constraints on population parameters., Comment: 23 pages, 13 figures. Version accepted for publication in Physical Review D

We search for strongly lensed and multiply imaged gravitational wave signals in the second observing run of Advanced LIGO and Advanced Virgo (O2). We exploit a new source of information, the so-called Morse phase, which further mitigates the search background and constrains viable lenses. The best candidate we find is consistent with a strongly lensed signal from a massive binary black hole (BBH) merger, with three detected images consisting of the previously catalogued events GW170104 and GW170814, and a subthreshold trigger, GWC170620. Given the number of BBH events detected so far, we estimate an overall false alarm probability $\sim 10^{-4}$ for the observed high degree of parameter coincidence between the three events. On the flip side, we measure the Morse phase differences which suggest a complex and atypical lens system, with at least five images including a magnified image at a local maximum of the Fermat potential. The low prior probability for multiple lensed images and the amount of fine tuning required in the lens model reduce the credibility of the lensing hypothesis. The long time delays between lensed images point toward a galaxy cluster lens with an internal velocity dispersion $\sigma \sim 650\,{\rm km/s}$, and the observed strain amplitudes imply a likely range $0.4 < z \lesssim 0.7$ for the source redshift. We provide an error ellipse of $\sim 16\,{\rm deg}^2$ for the sky location of the source together with additional specific constraints on the lens-host system, and encourage follow-up efforts to confirm or rule out any viable lens. If this is indeed a lensed event, successfully pinpointing the system would offer a unique opportunity to identify the host galaxy of a BBH merger, and even localize the source within it., Comment: 15 pages including references, 8 figures, and one table

Recent advances in Generative Adversarial Networks (GANs) have led to the creation of realistic-looking digital images that pose a major challenge to their detection by humans or computers. GANs are used in a wide range of tasks, from modifying small attributes of an image (StarGAN [14]), transferring attributes between image pairs (CycleGAN [91]), as well as generating entirely new images (ProGAN [36], StyleGAN [37], SPADE/GauGAN [64]). In this paper, we propose a novel approach to detect, attribute and localize GAN generated images that combines image features with deep learning methods. For every image, co-occurrence matrices are computed on neighborhood pixels of RGB channels in different directions (horizontal, vertical and diagonal). A deep learning network is then trained on these features to detect, attribute and localize these GAN generated/manipulated images. A large scale evaluation of our approach on 5 GAN datasets comprising over 2.76 million images (ProGAN, StarGAN, CycleGAN, StyleGAN and SPADE/GauGAN) shows promising results in detecting GAN generated images.

We have developed a perturbative method to model the resonant ionization of atomic systems in fluctuating laser fields. The perturbative method is based on an expansion in terms of the multitime cumulants, a suitable combination of moments (field's coherence functions), used to represent the field's statistical properties. The second-order truncated expansion is expressed in terms of the radiation's power spectrum and the intensity autocorrelation function. We investigate the range of validity of the model in terms of the field's coherence temporal length and peak intensity and have compared the results with conventional Monte-Carlo calculations. We apply the theory in the case of a near-resonant ionization of the Helium 2s2p autoionizing state with a SASE FEL pulse with square-exponentially dependent 1st-order coherence function. The ionization lineshape profile acquires a Voight profile; the degree of the Gaussian or Lorentzian character of which to depend crucially on the field's coherence time.

We perform a detailed parameter estimation study of binary black hole merger events reported in Zackay et al. and Venumadhav et al.. These are some of the faintest signals reported so far, and hence, relative to the loud events in the GWTC-1 catalog, the data should have lesser constraining power on their intrinsic parameters. Hence we examine the robustness of parameter inference to choices made in the analysis, as well as any potential systematics. We check the impact of different methods of estimating the noise power spectral density, different waveform models, and different priors for the compact object spins. For most of the events, the resulting differences in the inferred values of the parameters are much smaller than their statistical uncertainties. The estimation of the effective spin parameter $\chi_{\mathrm{eff}}$, i.e. the projection of the mass-weighted total spin along the angular momentum, can be sensitive to analysis choices for two of the sources with the largest effective spin magnitudes, GW151216 and GW170403. The primary differences arise from using a 3D isotropic spin prior: the tails of the posterior distributions should be interpreted with care and due consideration of the other data analysis choices.

Tumblr, as a leading content provider and social media, attracts 371 million monthly visits, 280 million blogs and 53.3 million daily posts. The popularity of Tumblr provides great opportunities for advertisers to promote their products through sponsored posts. However, it is a challenging task to target specific demographic groups for ads, since Tumblr does not require user information like gender and ages during their registration. Hence, to promote ad targeting, it is essential to predict user's demography using rich content such as posts, images and social connections. In this paper, we propose graph based and deep learning models for age and gender predictions, which take into account user activities and content features. For graph based models, we come up with two approaches, network embedding and label propagation, to generate connection features as well as directly infer user's demography. For deep learning models, we leverage convolutional neural network (CNN) and multilayer perceptron (MLP) to prediction users' age and gender. Experimental results on real Tumblr daily dataset, with hundreds of millions of active users and billions of following relations, demonstrate that our approaches significantly outperform the baseline model, by improving the accuracy relatively by 81% for age, and the AUC and accuracy by 5\% for gender.

We study the highly magnified arc SGAS J122651.3+215220 caused by a star-forming galaxy at $z_s=2.93$ crossing the lensing caustic cast by the galaxy cluster SDSS J1226+2152 ($z_l=0.43$), using Hubble Space Telescope observations. We report in the arc several asymmetric surface brightness features whose angular separations are a fraction of an arcsecond from the lensing critical curve and appear to be highly but unequally magnified image pairs of underlying compact sources, with one brightest pair having clear asymmetry consistently across four filters. One explanation of unequal magnification is microlensing by intracluster stars, which induces independent flux variations in the images of individual or groups of source stars in the lensed galaxy. For a second possibility, intracluster dark matter subhalos invisible to telescopes effectively perturb lensing magnifications near the critical curve and give rise to persistently unequal image pairs. Our modeling suggests, at least for the most prominent identified image pair, that the microlensing hypothesis is in tension with the absence of notable asymmetry variation over a six-year baseline, while subhalos of $\sim 10^6$--$10^8\,M_\odot$ anticipated from structure formation with Cold Dark Matter typically produce stationary and sizable asymmetries. We judge that observations at additional times and more precise lens models are necessary to stringently constrain temporal variability and robustly distinguish between the two explanations. The arc under this study is a scheduled target of a Director's Discretionary Early Release Science program of the James Webb Space Telescope, which will provide deep images and a high-resolution view with integral field spectroscopy., Comment: New version accepted by MNRAS; 18 pages including references and appendices, 13 figures and 4 tables; major revision of Sec. 3.2 and Figure 4 presenting improved data analysis; original conclusion strengthened

We introduce a new technique to search for gravitational wave events from compact binary mergers that produce a clear signal only in a single gravitational wave detector, and marginal signals in other detectors. Such a situation can arise when the detectors in a network have different sensitivities, or when sources have unfavorable sky locations or orientations. We start with a short list of loud single-detector triggers from regions of parameter space that are empirically unaffected by glitches (after applying signal-quality vetoes). For each of these triggers, we compute evidence for astrophysical origin from the rest of the detector network by coherently combining the likelihoods from all detectors and marginalizing over extrinsic geometric parameters. We report the discovery of two new binary black hole (BBH) mergers in the second observing run of Advanced LIGO and Virgo (O2), in addition to the ones that were reported in Abbott et al. (2018) and Venumadhav et al. (2019). We estimate that the two events have false alarm rates of one in 19 years (60 O2) and one in 11 years (36 O2). One of the events, GW170817A, has primary and secondary masses $m_1^{\rm src} = 56_{-10}^{+16} \, M_\odot$ and $m_2^{\rm src} = 40_{-11}^{+10} \, M_\odot$ in the source frame. The existence of GW170817A should be very informative about the theoretically predicted upper mass gap for stellar mass black holes. Its effective spin parameter is measured to be $\chi_{\rm eff} = 0.5 \pm 0.2$, which is consistent with the tendency of the heavier detected BBH systems to have large and positive effective spin parameters. The other event, GWC170402, will be discussed thoroughly in future work., Comment: Comments are welcome

Searches for gravitational waves crucially depend on exact signal processing of noisy strain data from gravitational wave detectors, which are known to exhibit significant non-Gaussian behavior. In this paper, we study two distinct non-Gaussian effects in the LIGO/Virgo data which reduce the sensitivity of searches: first, variations in the noise power spectral density (PSD) on timescales of more than a few seconds; and second, loud and abrupt transient `glitches' of terrestrial or instrumental origin. We derive a simple procedure to correct, at first order, the effect of the variation in the PSD on the search background. Given the knowledge of the existence of localized glitches in particular segments of data, we also develop a method to insulate statistical inference from these glitches, so as to cleanly excise them without affecting the search background in neighboring seconds. We show the importance of applying these methods on the publicly available LIGO data, and measure an increase in the detection volume of at least $15\%$ from the PSD-drift correction alone, due to the improved background distribution., Comment: comments welcome