Searcharxiv⌕ Search

arXiv subjects

Cheng Li

Publications and source records attributed to Cheng Li.

At least 361 records · Page 20Linked to original sources
arXiv2020

SDSS-IV MaNGA: Bayesian analysis of the star formation history of low-mass galaxies in the local Universe

We measure the star formation histories (SFH) of a sample of low-mass galaxies with $M_\ast<10^9M_\odot$ from the SDSS-IV MaNGA survey. The large number of IFU spectra for each galaxy are either combined to reach a high signal to noise ratio or used to investigate spatial variations. We use Bayesian inferences based on full spectrum fitting. Our analysis based on Bayesian evidence ratio indicates a strong preferencefor a model that allows the presence of an old stellar population, and that an improper model for the SFH can significantly underestimate the old population in these galaxies. The addition of NIR photometry to the constraining data can further distinguish between different SFH model families and significantly tighten the constraints on the mass fraction in the old population. On average more than half of the stellar mass in present-day low-mass galaxies formed at least 8 Gyrs ago, while about 30\% within the past 4 Gyrs. Satellite galaxies on average have formed their stellar mass earlier than central galaxies. The radial dependence of the SFH is quite weak. Our results suggest that most of the low-mass galaxies have an early episode of active star formation that produces a large fraction of their present stellar mass.

astro-ph.GA↗
arXiv2020

Study of the water Cherenkov detector with high dynamic range for LHAASO

The Large High-Altitude Air Shower Observatory (LHAASO) is being built at Haizi mountain, Sichuan, China at an altitude of 4410 m. One of its main goals is to survey the northern sky for very-high-energy (above 100 GeV) gamma ray sources via its ground-based Water Cherenkov Detector Array (WCDA). WCDA is 78000 $m^2$ in dimension and consists of 3120 water detector cells divided into 3 water ponds. A hemispherical 8-inch photomultiplier tube (PMT) CR365-02-1 from Beijing Hamamatsu Photon Techniques INC. (BHP) is installed at the bottom-center of each cell of the first water pond to collect Cherenkov light produced by air shower particles crossing water. This proceeding includes the technical design of WCDA, the design of a high dynamic range base for CR365-02-1, the PMT test system and test results of 997 PMTs.

physics.ins-det↗
arXiv2020

SDSS-IV MaNGA: the indispensable role of bars in enhancing the central star formation of low-$z$ galaxies

We analyse two-dimensional maps and radial profiles of EW(H$α$), EW(H$δ_A$), and D$_n$(4000) of low-redshift galaxies using integral field spectroscopy from the MaNGA survey. Out of $\approx1400$ nearly face-on late-type galaxies with a redshift $z<0.05$, we identify 121 "turnover" galaxies that each have a central upturn in EW(H$α$), EW(H$δ_A$) and/or a central drop in D$_n$(4000), indicative of ongoing/recent star formation. The turnover features are found mostly in galaxies with a stellar mass above $\sim$10$^{10}$ M$_{\odot}$ and NUV-$r$ colour less than $\approx5$. The majority of the turnover galaxies are barred, with a bar fraction of 89$\pm$3\%. Furthermore, for barred galaxies the radius of the central turnover region is found to tightly correlate with one third of the bar length. Comparing the observed and the inward extrapolated star formation rate surface density, we estimate that the central SFR have been enhanced by an order of magnitude. Conversely, only half of the barred galaxies in our sample have a central turnover feature, implying that the presence of a bar is not sufficient to lead to a central SF enhancement. We further examined the SF enhancement in paired galaxies, as well as the local environment, finding no relation. This implies that environment is not a driving factor for central SF enhancement in our sample. Our results reinforce both previous findings and theoretical expectation that galactic bars play a crucial role in the secular evolution of galaxies by driving gas inflow and enhancing the star formation and bulge growth in the center.

astro-ph.GA↗
arXiv2020

A new estimator of resolved molecular gas in nearby galaxies

A relationship between dust-reprocessed light from recent star formation and the amount of star-forming gas in a galaxy produces a correlation between WISE 12 $μ$m emission and CO line emission. Here we explore this correlation on kiloparsec scales with CO(1-0) maps from EDGE-CALIFA matched in resolution to WISE 12 $μ$m images. We find strong CO-12 $μ$m correlations within each galaxy and we show that the scatter in the global CO-12 $μ$m correlation is largely driven by differences from galaxy to galaxy. The correlation is stronger than that between star formation rate and H$_2$ surface densities ($Σ(\mathrm{H_2})$). We explore multi-variable regression to predict $Σ(\mathrm{H_2})$ in star-forming pixels using the WISE 12 $μ$m data combined with global and resolved galaxy properties, and provide the fit parameters for the best estimators. We find that $Σ(\mathrm{H_2})$ estimators that include $Σ(\mathrm{12\>μm})$ are able to predict $Σ(\mathrm{H_2})$ more accurately than estimators that include resolved optical properties instead of $Σ(\mathrm{12\>μm})$. These results suggest that 12 $μ$m emission and H$_2$ as traced by CO emission are physically connected at kiloparsec scales. This may be due to a connection between polycyclic aromatic hydrocarbon (PAH) emission and the presence of H$_2$. The best single-property estimator is $\log \frac{Σ(\mathrm{H_2})}{\mathrm{M_\odot\>pc^{-2}}} = (0.48 \pm 0.01) + (0.71 \pm 0.01)\log \frac{Σ(\mathrm{12\>μm})}{\mathrm{L_\odot\>pc^{-2}}}$. This correlation can be used to efficiently estimate $Σ(\mathrm{H_2})$ down to at least $1 \> M_\odot \> \mathrm{pc^{-2}}$ in star-forming regions within nearby galaxies.

astro-ph.GA↗
arXiv2020

SDSS-IV MaNGA: The Nature of an Off-galaxy H$α$ Blob -- A Multi-wavelength View of Offset Cooling in a Merging Galaxy Group

Galaxies in dense environments, such as groups and clusters, experience various processes by which galaxies gain and lose gas. Using data from the SDSS-IV MaNGA survey, we previously reported the discovery of a giant (6 -- 8 kpc in diameter) H$α$ blob, Totoro, about 8 kpc away from a pair of galaxies (Satsuki and Mei) residing in a galaxy group which is experiencing a group-group merger. Here, we combine interferometric $^{12}$CO(1--0) molecular gas data, new wide-field H$α$, $u$-band data, and published X-ray data to determine the origin of the blob. Several scenarios are discussed to account for its multi-wavelength properties, including (1) H$α$ gas being stripped from galaxy Satsuki by ram-pressure; (2) a separated low-surface-brightness galaxy; (3) gas being ejected or ionized by an active galactic nucleus (AGN); and (4) a cooling intra-group medium (IGM). Scenarios (1) and (2) are less favored by the present data. Scenario (3) is also less likely as there is no evidence for an active ongoing AGN in the host galaxy. We find that the CO (cold) and H$α$ (warm) gas coexist with X-ray (hot) structures; moreover, the derived cooling time is within the regime where molecular and H$α$ gas are expected. The coexistence of gas with different temperatures also agrees with that of cooling gas in other systems. Our multi-wavelength results strongly suggest that the CO and H$α$ gas are the product of cooling from the IGM at its current location, i.e., cooling has occurred, and may be ongoing, well outside the host-galaxy core.

astro-ph.GA↗
arXiv2020

Constraints on circum-galactic media from Sunyaev-Zel'dovich effects and X-ray data

We use observational measurements of thermal and kinetic Sunyaev-Zel'dovich effects, as well as soft X-ray emission associated with galaxy groups to constrain the gas density and temperature in the circumgalactic medium (CGM) for dark matter halos with masses above $10^{12.5}M_{\odot}$. A number of generic models are used together with a Bayesian scheme to make model inferences. We find that gas with a single temperature component cannot provide a consistent model to match the observational data. A simple two-phase model assuming a hot component and an ionized warm component can accommodate all the three observations. The total amount of the gas in individual halos is inferred to be comparable to the universal baryon fraction corresponding to the halo mass. The inferred temperature of the hot component is comparable to the halo virial temperature. The fraction of the hot component increases from $(15-40)\%$ for $10^{12.5}{M}_\odot$ halos to $(40-60)\%$ for $10^{14.5}{M}_\odot$ halos, where the ranges reflect uncertainties in the assumed gas density profile. Our results suggest that a significant fraction of the halo gas is in a non-thermalized component with temperature much lower than the virial temperature.

astro-ph.GA↗
arXiv2020

Identifying galaxy groups at high redshift from incomplete spectroscopic data: I. The group finder and application to zCOSMOS

Identifying galaxy groups from redshift surveys of galaxies plays an important role in connecting galaxies with the underlying dark matter distribution. Current and future high-$z$ spectroscopic surveys, usually incomplete in redshift sampling, present both opportunities and challenges to identifying groups in the high-$z$ Universe. We develop a group finder that is based on incomplete redshift samples combined with photometric data, using a machine learning method to assign halo masses to identified groups. Test using realistic mock catalogs shows that $\gtrsim 90\%$ of true groups with halo masses $\rm M_h \gtrsim 10^{12} M_{\odot}/h$ are successfully identified, and that the fraction of contaminants is smaller than $10\%$. The standard deviation in the halo mass estimation is smaller than 0.25 dex at all masses. We apply our group finder to zCOSMOS-bright and describe basic properties of the group catalog obtained.

astro-ph.GA↗
arXiv2020

Simultaneously-Collected Multimodal Lying Pose Dataset: Towards In-Bed Human Pose Monitoring under Adverse Vision Conditions

Computer vision (CV) has achieved great success in interpreting semantic meanings from images, yet CV algorithms can be brittle for tasks with adverse vision conditions and the ones suffering from data/label pair limitation. One of this tasks is in-bed human pose estimation, which has significant values in many healthcare applications. In-bed pose monitoring in natural settings could involve complete darkness or full occlusion. Furthermore, the lack of publicly available in-bed pose datasets hinders the use of many successful pose estimation algorithms for this task. In this paper, we introduce our Simultaneously-collected multimodal Lying Pose (SLP) dataset, which includes in-bed pose images from 109 participants captured using multiple imaging modalities including RGB, long wave infrared, depth, and pressure map. We also present a physical hyper parameter tuning strategy for ground truth pose label generation under extreme conditions such as lights off and being fully covered by a sheet/blanket. SLP design is compatible with the mainstream human pose datasets, therefore, the state-of-the-art 2D pose estimation models can be trained effectively with SLP data with promising performance as high as 95% at PCKh@0.5 on a single modality. The pose estimation performance can be further improved by including additional modalities through collaboration.

cs.CV↗
arXiv2020

P-Cloth: Interactive Complex Cloth Simulation on Multi-GPU Systems using Dynamic Matrix Assembly and Pipelined Implicit Integrators

We present a novel parallel algorithm for cloth simulation that exploits multiple GPUs for fast computation and the handling of very high resolution meshes. To accelerate implicit integration, we describe new parallel algorithms for sparse matrix-vector multiplication (SpMV) and for dynamic matrix assembly on a multi-GPU workstation. Our algorithms use a novel work queue generation scheme for a fat-tree GPU interconnect topology. Furthermore, we present a novel collision handling scheme that uses spatial hashing for discrete and continuous collision detection along with a non-linear impact zone solver. Our parallel schemes can distribute the computation and storage overhead among multiple GPUs and enable us to perform almost interactive simulation on complex cloth meshes, which can hardly be handled on a single GPU due to memory limitations. We have evaluated the performance with two multi-GPU workstations (with 4 and 8 GPUs, respectively) on cloth meshes with 0.5-1.65M triangles. Our approach can reliably handle the collisions and generate vivid wrinkles and folds at 2-5 fps, which is significantly faster than prior cloth simulation systems. We observe almost linear speedups with respect to the number of GPUs.

cs.GR↗
arXiv2020

Relating the structure of dark matter halos to their assembly and environment

We use a large $N$-body simulation to study the relation of the structural properties of dark matter halos to their assembly history and environment. The complexity of individual halo assembly histories can be well described by a small number of principal components (PCs), which, compared to formation times, provide a more complete description of halo assembly histories and have a stronger correlation with halo structural properties. Using decision trees built with the random ensemble method, we find that about $60\%$, $10\%$, and $20\%$ of the variances in halo concentration, axis ratio, and spin, respectively, can be explained by combining four dominating predictors: the first PC of the assembly history, halo mass, and two environment parameters. Halo concentration is dominated by halo assembly. The local environment is found to be important for the axis ratio and spin but is degenerate with halo assembly. The small percentages of the variance in the axis ratio and spin that are explained by known assembly and environmental factors suggest that the variance is produced by many nuanced factors and should be modeled as such. The relations between halo intrinsic properties and environment are weak compared to their variances, with the anisotropy of the local tidal field having the strongest correlation with halo properties. Our method of dimension reduction and regression can help simplify the characterization of the halo population and clarify the degeneracy among halo properties.

astro-ph.GA↗
arXiv2020

Towards in-store multi-person tracking using head detection and track heatmaps

Computer vision algorithms are being implemented across a breadth of industries to enable technological innovations. In this paper, we study the problem of computer vision based customer tracking in retail industry. To this end, we introduce a dataset collected from a camera in an office environment where participants mimic various behaviors of customers in a supermarket. In addition, we describe an illustrative example of the use of this dataset for tracking participants based on a head tracking model in an effort to minimize errors due to occlusion. Furthermore, we propose a model for recognizing customers and staff based on their movement patterns. The model is evaluated using a real-world dataset collected in a supermarket over a 24-hour period that achieves 98% accuracy during training and 93% accuracy during evaluation.

cs.CV↗
arXiv2020

A Global Non-Hydrostatic Atmospheric Model with a Mass and Energy Conserving Vertically-Implicit-Correction (VIC) Scheme

Global non-hydrostatic atmospheric models are becoming increasingly important for studying the climates of planets and exoplanets. However, such models suffer from computational difficulties due to the large aspect ratio between the horizontal and vertical directions. To overcome this problem, we developed a global model using a vertically-implicit-correction (VIC) scheme in which the integration time step is no longer limited by the propagation of acoustic waves in the vertical. We proved that our model, based on the $\rm Athena^{++}$ framework and its extension for planetary atmospheres - SNAP (Simulating Non-hydrostatic Atmosphere on Planets), rigorously conserves mass and energy in finite volume simulations. We found that traditional numerical stabilizers such as hyper-viscosity and divergence damping are not needed when using the VIC scheme, which greatly simplifies the numerical implementation and improves stability. We present simulation results ranging from 1D linear waves to 3D global circulations with and without the VIC scheme. These tests demonstrate that our formulation correctly tracks local turbulent motions, produces Kelvin-Helmholtz instability, and generates a super-rotating jet on hot Jupiters. Employing this VIC scheme improves the computational efficiency of global simulations by more than two orders of magnitude compared to an explicit model and facilitates the capability of simulating a wide range of planetary atmospheres both regionally and globally.

astro-ph.EP↗
arXiv2020

Settling and Clustering of Snow Particles in Atmospheric Turbulence

The effect of turbulence on snow precipitation is not incorporated into present weather forecasting models. Here we show evidence that turbulence is in fact a key influence on both fall speed and spatial distribution of settling snow. We consider three snowfall events under vastly different levels of atmospheric turbulence. We characterize the size and morphology of the snow particles, and we simultaneously image their velocity, acceleration, and relative concentration over vertical planes about 30 m2 in area. We find that turbulence-driven settling enhancement explains otherwise contradictory trends between the particle size and velocity. The estimates of the Stokes number and the correlation between vertical velocity and local concentration indicate that the enhanced settling is rooted in the preferential sweeping mechanism. When the snow vertical velocity is large compared to the characteristic turbulence velocity, the crossing trajectories effect results in strong accelerations. When the conditions of preferential sweeping are met, the concentration field is highly non-uniform and clustering appears over a wide range of scales. These clusters, identified for the first time in a naturally occurring flow, display the signature features seen in canonical settings: power-law size distribution, fractal-like shape, vertical elongation, and large fall speed that increases with the cluster size. These findings demonstrate that the fundamental phenomenology of particle-laden turbulence can be leveraged towards a better predictive understanding of snow precipitation and ground snow accumulation. They also demonstrate how environmental flows can be used to investigate dispersed multiphase flows at Reynolds numbers not accessible in laboratory experiments or numerical simulations.

physics.flu-dyn↗
arXiv2020

Sparse Spectrum Gaussian Process for Bayesian Optimization

We propose a novel sparse spectrum approximation of Gaussian process (GP) tailored for Bayesian optimization. Whilst the current sparse spectrum methods provide desired approximations for regression problems, it is observed that this particular form of sparse approximations generates an overconfident GP, i.e. it produces less epistemic uncertainty than the original GP. Since the balance between predictive mean and the predictive variance is the key determinant to the success of Bayesian optimization, the current sparse spectrum methods are less suitable for it. We derive a new regularized marginal likelihood for finding the optimal frequencies to fix this over-confidence issue, particularly for Bayesian optimization. The regularizer trades off the accuracy in the model fitting with a targeted increase in the predictive variance of the resultant GP. Specifically, we use the entropy of the global maximum distribution from the posterior GP as the regularizer that needs to be maximized. Since this distribution cannot be calculated analytically, we first propose a Thompson sampling based approach and then a more efficient sequential Monte Carlo based approach to estimate it. Later, we also show that the Expected Improvement acquisition function can be used as a proxy for the maximum distribution, thus making the whole process further efficient. Experiments show considerable improvement to Bayesian optimization convergence rate over the vanilla sparse spectrum method and over a full GP when its covariance matrix is ill-conditioned due to the presence of a large number of observations.

cs.LG↗
arXiv2020

XSP: Across-Stack Profiling and Analysis of Machine Learning Models on GPUs

There has been a rapid proliferation of machine learning/deep learning (ML) models and wide adoption of them in many application domains. This has made profiling and characterization of ML model performance an increasingly pressing task for both hardware designers and system providers, as they would like to offer the best possible system to serve ML models with the target latency, throughput, cost, and energy requirements while maximizing resource utilization. Such an endeavor is challenging as the characteristics of an ML model depend on the interplay between the model, framework, system libraries, and the hardware (or the HW/SW stack). Existing profiling tools are disjoint, however, and only focus on profiling within a particular level of the stack, which limits the thoroughness and usefulness of the profiling results. This paper proposes XSP - an across-stack profiling design that gives a holistic and hierarchical view of ML model execution. XSP leverages distributed tracing to aggregate and correlates profile data from different sources. XSP introduces a leveled and iterative measurement approach that accurately captures the latencies at all levels of the HW/SW stack in spite of the profiling overhead. We couple the profiling design with an automated analysis pipeline to systematically analyze 65 state-of-the-art ML models. We demonstrate that XSP provides insights which would be difficult to discern otherwise.

cs.LG↗
arXiv2020

A Complex KBQA System using Multiple Reasoning Paths

Multi-hop knowledge based question answering (KBQA) is a complex task for natural language understanding. Many KBQA approaches have been proposed in recent years, and most of them are trained based on labeled reasoning path. This hinders the system's performance as many correct reasoning paths are not labeled as ground truth, and thus they cannot be learned. In this paper, we introduce an end-to-end KBQA system which can leverage multiple reasoning paths' information and only requires labeled answer as supervision. We conduct experiments on several benchmark datasets containing both single-hop simple questions as well as muti-hop complex questions, including WebQuestionSP (WQSP), ComplexWebQuestion-1.1 (CWQ), and PathQuestion-Large (PQL), and demonstrate strong performance.

cs.LG↗
arXiv2020

Wolf-Rayet galaxies in SDSS-IV MaNGA. I. Catalog construction and sample properties

Wolf-Rayet (WR) galaxies are a rare population of galaxies that host living high-mass stars during their WR phase (i.e. WR stars) and are thus expected to provide interesting constraints on the stellar Initial Mass Function, massive star formation, stellar evolution models, etc. Spatially resolved spectroscopy should in principle provide a more efficient way of identifying WR galaxies than single-fiber surveys of galactic centers such as SDSS-I & II, as WR stars should be more preferentially found in discs. Using IFU data from the ongoing SDSS-IV MaNGA survey, we have performed a thorough search for WR galaxies. We first identify H II regions in each datacube and carry out full spectral fitting to the stacked spectra. We then visually inspect the residual spectrum of each H II region and identify WR regions that present a significant "blue bump" at 4600-4750 A. The resulting WR catalog includes 267 WR regions of ~500pc (radius) sizes, distributed in 90 galaxies from the current sample of MaNGA (MaNGA Product Launch 7). We find WR regions are exclusively found in galaxies that show bluest colors and highest star formation rates for their mass. Most WR galaxies have late-type morphologies and show relatively large asymmetry in their images, implying that WR regions are more preferentially found in interacting/merging galaxies. We estimate the stellar mass function of WR galaxies and the mass-dependent detection rate. The detection rate of WR galaxies is typically ~2%, with weak dependence on stellar mass. This detection rate is about 40 times higher than previous studies with SDSS single fiber data, and by a factor of 2 lower than the CALIFA-based WR catalog. We make comparisons with SDSS and CALIFA studies, and conclude that different detection rates can be explained mainly by three factors: spatial coverage, spectral signal-to-noise ratio, and redshift ranges of the parent sample.

astro-ph.GA↗