BZPEERReading, trust and collaboration for research
Menu

Discover

GraphFollow connections around a paper, topic or saved view.

Workspaces

HomePick up where your research flow left off.InboxHandle requests, replies and notifications from one place.CollectionsCurate reading lists, evidence sets and shareable dossiers.ResearchSee your private provenance graph, trail and imports.CollaborationMove from discovery into focused discussion rooms.PublishDraft, preview and publish full-text research articles.

Network

ExpertsFind specialists worth following or asking for help.CommunitiesJoin research circles organized around shared work.PartnersSee external organizations active around the same topics.

Opportunities

ListingsBrowse roles, calls and collaborations with clearer fit signals.

Account

Log inReturn to your reading and collaboration workspace.Create accountStart saving work, following people and joining teams.
Log inCreate account

Researcher profile

Guilin Liu

Guilin Liu contributes to research discovery and scholarly infrastructure.

ResearcherAffiliation not importedOpen to collaborate
astro-ph.GAComputer VisionGraphicsastro-ph.CO

Trust snapshot

Quick read

Trust 21 - EmergingVerification L1Unclaimed author
9works
0followers
4topics
4close collaborators

Actions

Decide how to stay connected

Follow researcher0

Identity and collaboration

How to connect with this researcher

Claiming links this public author record to a researcher profile and unlocks direct collaboration workflows.

Log in to claim

Direct collaboration

Open a focused conversation when the fit is right

Claim this author entity first to unlock direct invitations.

Research graph

See the researcher in context

Open full explorer

Inspect adjacent work, topics, institutions and collaborators without jumping out to a separate graph page.

Building this graph slice

BZPEER is loading the nearby papers, people, topics and institutions for this page.

Published work

9 published item(s)

preprint2022arXiv

CO Excitation and its Connection to Star Formation at 200 pc in NGC 1365

We report high resolution 2" ~ 200 pc mappings of the central region of the nearby barred spiral galaxy NGC 1365 in the CO(1--0) and CO(2--1) emission lines. The 2--1/1--0 ratio of integrated intensities shows a large scatter (0.15) with a median value of 0.67. We also calculate the ratio of velocity dispersions and peak temperatures and find that in most cases the velocity dispersion ratio is close to unity and thus the peak temperature ratio is comparable to the integrated intensity ratio. This result indicates that both CO(1--0) and CO(2--1) lines trace similar components of molecular gas, with their integrated intensity (or peak temperature) ratios reflecting the gas density and/or temperature. Similar to recent kpc scale studies, these ratios show a positive correlation with a star formation rate indicator (here we use an extinction-corrected H-alpha map), suggesting that molecular gas associated with recent star formation is denser and/or warmer. We also find that some CO spectra show two peaks owing to complicated kinematics, and such two components likely trace molecular gas at different conditions. This result demonstrates the importance of spectral fitting to measure integrated intensities and their ratios more accurately.

0 citations0 reviewsNo code linkedOpen access
preprint2022arXiv

Dual Contrastive Loss and Attention for GANs

Generative Adversarial Networks (GANs) produce impressive results on unconditional image generation when powered with large-scale image datasets. Yet generated images are still easy to spot especially on datasets with high variance (e.g. bedroom, church). In this paper, we propose various improvements to further push the boundaries in image generation. Specifically, we propose a novel dual contrastive loss and show that, with this loss, discriminator learns more generalized and distinguishable representations to incentivize generation. In addition, we revisit attention and extensively experiment with different attention blocks in the generator. We find attention to be still an important module for successful image generation even though it was not used in the recent state-of-the-art models. Lastly, we study different attention architectures in the discriminator, and propose a reference attention mechanism. By combining the strengths of these remedies, we improve the compelling state-of-the-art Fréchet Inception Distance (FID) by at least 17.5% on several benchmark datasets. We obtain even more significant improvements on compositional synthetic scenes (up to 47.5% in FID). Code and models are available at https://github.com/ningyu1991/AttentionDualContrastGAN .

0 citations0 reviewsNo code linkedOpen access
preprint2022arXiv

Evidence for quasar fast outflows being accelerated at the scale of tens of parsecs

Quasar outflows may play a crucial role in regulating the host galaxy, although the spatial scale of quasar outflows remain a major enigma, with their acceleration mechanism poorly understood. The kinematic information of outflow is the key to understanding its origin and acceleration mechanism. Here, we report the galactocentric distances of different outflow components for both a sample and an individual quasar. We find that the outflow distance increases with velocity, with a typical value from several parsecs to more than one hundred parsecs, providing direct evidence for an acceleration happening at a scale of the order of 10 parsecs. These outflows carry ~1% of the total quasar energy, while their kinematics are consistent with a dust driven model with a launching radius comparable to the scale of a dusty torus, indicating that the coupling between dust and quasar radiation may produce powerful feedback that is crucial to galaxy evolution.

0 citations0 reviewsNo code linkedOpen access
preprint2022arXiv

The ALPINE-ALMA [CII] survey: The infrared-radio correlation and AGN fraction of star-forming galaxies at z $\sim$ 4.4-5.9

We present the radio properties of 66 spectroscopically-confirmed normal star-forming galaxies (SFGs) at $4.4<z<5.9$ in the COSMOS field that were [C II] detected in the Atacama Large Millimeter Array (ALMA) Large Program to INvestigate [C II] at Early times (ALPINE). We separate these galaxies (&#34;CII-detected-all&#34;) into lower redshift (&#34;CII-detected-lz&#34;, $\langle z\rangle=4.5$) and higher redshift (&#34;CII-detected-hz&#34;, $\langle z\rangle=5.6$) sub-samples and stack multi-wavelength imaging for each sub-sample from X-ray to radio bands. A radio signal is detected in the stacked 3 GHz image of CII-detected-all and -lz samples at $\gtrsim3σ$. We find that the infrared-radio correlation of our sample, quantified by $q_{\mathrm{TIR}}$, is lower than the local relation for normal SFGs at $\sim$3$σ$ significance level, and is instead broadly consistent with that of bright sub-mm galaxies at $2<z<5$. Neither of these samples show evidence of dominant AGN activity in their stacked Spectral Energy Distributions (SEDs), rest-frame UV spectra, or X-ray images. Although we cannot rule out the possible effect of the assumed spectral index and the applied infrared SED templates as at least partially causing these differences, the lower obscured fraction of star formation than at lower redshift can alleviate the tension between our stacked $q_{\mathrm{TIR}}$s and that of local normal SFGs. It is possible that the dust buildup, which primarily governs the IR emission in addition to older stellar populations, has not had enough time to occur fully in these galaxies, whereas the radio emission can respond on a more rapid timescale. Therefore, we might expect a lower $q_{\mathrm{TIR}}$ to be a general property of high-redshift SFGs.

0 citations0 reviewsNo code linkedOpen access
preprint2021arXiv

An Extraordinary Response of Iron Emission to the Central Outburst in a Tidal Disruption Event Candidate

Understanding the origin of \feii\ emission is important because it is crucial to construct the main sequence of Active Galactic Nuclei (AGNs). Despite several decades of observational and theoretical effort, the location of the optical iron emitting region and the mechanism responsible for the positive correlation between the \feii\ strength and the black hole accretion rate remain open questions as yet. In this letter, we report the optical \feii\ response to the central outburst in PS1-10adi, a candidate tidal disruption event (TDE) taking place in an AGN at $z = 0.203$ that has aroused extensive attention. For the first time, we observe that the \feii\ response in the rising phase of its central luminosity is significantly more prominent than that in the decline phase, showing a hysteresis effect. We interpret this hysteresis effect as a consequence of the gradual sublimation of the dust grains situating at the inner surface of the torus into gas when the luminosity of the central engine increases. It is the iron element released from the sublimated dust that contributes evidently to the observed \feii\ emission. This interpretation, together with the weak response of the \hb\ emission as we observe, naturally explains the applicability of relative \feii\ strength as a tracer of the Eddington ratio. In addition, optical iron emission of this origin renders the \feii\ time lag a potential &#34;standard candle&#34; with cosmological implications.

0 citations0 reviewsNo code linkedOpen access
preprint2021arXiv

Evidence for the connection between star formation rate and evolutionary phases of quasars

Both theory and observations suggest that outflows driven by an active central supermassive black hole (SMBH) has a feedback effect on shaping the global properties of the host galaxy. However, whether feedback from the outflow is effective, and if so, whether it is positive or negative, has long been controversial. Here, using the latest catalog from the Sloan Digital Sky Survey (SDSS), we use the flux ratio of the [O II] to [Ne V] emission lines as a proxy to compare the star formation rate (SFR) in the hosts of quasars with different types of broad absorption lines (BALs): low-ionization (Lo)BAL, high-ionization (Hi)BAL, and non-BAL. We find that SFR decreases from LoBAL to HiBAL quasars, and then increases from HiBAL to non-BAL quasars. Assuming that the sequence of LoBAL to HiBAL to non-BAL represents evolution, our results are consistent with a quenching and subsequent rebound of star formation in quasar host galaxies. This phenomenon can be explained that the SFR is suppressed by the outflow, which then rebounds once the outflow disappears as the quasars evolve from HiBALs to non-BALs. Our result suggests that the quasar outflow has a negative global feedback on galaxy evolution.

0 citations0 reviewsNo code linkedOpen access
preprint2020arXiv

A sharp rise in the detection rate of broad absorption line variations in a quasar SDSS J141955.26+522741.1

We present an analysis of the variability of broad absorption lines (BALs) in a quasar SDSS J141955.26+522741.1 at $z=2.145$ with 72 observations from the Sloan Digital Sky Survey Data Release 16 (SDSS DR16). The strong correlation between the equivalent widths of BAL and the continuum luminosity, reveals that the variation of BAL trough is dominated by the photoionization. The photoionization model predicts that when the time interval $ΔT$ between two observations is longer than the recombination timescale $t_{\rm rec}$, the BAL variations can be detected. This can be characterized as a &#34;sharp rise&#34; in the detection rate of BAL variation at $ΔT=t_{\rm rec}$. For the first time, we detect such a &#34;sharp rise&#34; signature in the detection rate of BAL variations. As a result, we propose that the $t_{\rm rec}$ can be obtained from the &#34;sharp rise&#34; of the detection rate of BAL variation. It is worth mentioning that the BAL variations are detected at the time-intervals less than the $t_{\rm rec}$ for half an order of magnitude in two individual troughs. This result indicates that there may be multiple components with different $t_{\rm rec}$ but the same velocity in an individual trough.

0 citations0 reviewsNo code linkedOpen access
preprint2020arXiv

Panoptic-based Image Synthesis

Conditional image synthesis for generating photorealistic images serves various applications for content editing to content generation. Previous conditional image synthesis algorithms mostly rely on semantic maps, and often fail in complex environments where multiple instances occlude each other. We propose a panoptic aware image synthesis network to generate high fidelity and photorealistic images conditioned on panoptic maps which unify semantic and instance information. To achieve this, we efficiently use panoptic maps in convolution and upsampling layers. We show that with the proposed changes to the generator, we can improve on the previous state-of-the-art methods by generating images in complex instance interaction environments in higher fidelity and tiny objects in more details. Furthermore, our proposed method also outperforms the previous state-of-the-art methods in metrics of mean IoU (Intersection over Union), and detAP (Detection Average Precision).

0 citations0 reviewsNo code linkedOpen access
preprint2020arXiv

Transposer: Universal Texture Synthesis Using Feature Maps as Transposed Convolution Filter

Conventional CNNs for texture synthesis consist of a sequence of (de)-convolution and up/down-sampling layers, where each layer operates locally and lacks the ability to capture the long-term structural dependency required by texture synthesis. Thus, they often simply enlarge the input texture, rather than perform reasonable synthesis. As a compromise, many recent methods sacrifice generalizability by training and testing on the same single (or fixed set of) texture image(s), resulting in huge re-training time costs for unseen images. In this work, based on the discovery that the assembling/stitching operation in traditional texture synthesis is analogous to a transposed convolution operation, we propose a novel way of using transposed convolution operation. Specifically, we directly treat the whole encoded feature map of the input texture as transposed convolution filters and the features&#39; self-similarity map, which captures the auto-correlation information, as input to the transposed convolution. Such a design allows our framework, once trained, to be generalizable to perform synthesis of unseen textures with a single forward pass in nearly real-time. Our method achieves state-of-the-art texture synthesis quality based on various metrics. While self-similarity helps preserve the input textures&#39; regular structural patterns, our framework can also take random noise maps for irregular input textures instead of self-similarity maps as transposed convolution inputs. It allows to get more diverse results as well as generate arbitrarily large texture outputs by directly sampling large noise maps in a single pass as well.

0 citations0 reviewsNo code linkedOpen access