arXiv subjects

Yun Wang

Publications and source records attributed to Yun Wang.

At least 217 records · Page 12Linked to original sources

arXiv2022

OneLabeler: A Flexible System for Building Data Labeling Tools

Labeled datasets are essential for supervised machine learning. Various data labeling tools have been built to collect labels in different usage scenarios. However, developing labeling tools is time-consuming, costly, and expertise-demanding on software development. In this paper, we propose a conceptual framework for data labeling and OneLabeler based on the conceptual framework to support easy building of labeling tools for diverse usage scenarios. The framework consists of common modules and states in labeling tools summarized through coding of existing tools. OneLabeler supports configuration and composition of common software modules through visual programming to build data labeling tools. A module can be a human, machine, or mixed computation procedure in data labeling. We demonstrate the expressiveness and utility of the system through ten example labeling tools built with OneLabeler. A user study with developers provides evidence that OneLabeler supports efficient building of diverse data labeling tools.

cs.HC↗

arXiv2022

CT-SAT: Contextual Transformer for Sequential Audio Tagging

Sequential audio event tagging can provide not only the type information of audio events, but also the order information between events and the number of events that occur in an audio clip. Most previous works on audio event sequence analysis rely on connectionist temporal classification (CTC). However, CTC's conditional independence assumption prevents it from effectively learning correlations between diverse audio events. This paper first attempts to introduce Transformer into sequential audio tagging, since Transformers perform well in sequence-related tasks. To better utilize contextual information of audio event sequences, we draw on the idea of bidirectional recurrent neural networks, and propose a contextual Transformer (cTransformer) with a bidirectional decoder that could exploit the forward and backward information of event sequences. Experiments on the real-life polyphonic audio dataset show that, compared to CTC-based methods, the cTransformer can effectively combine the fine-grained acoustic representations from the encoder and coarse-grained audio event cues to exploit contextual information to successfully recognize and predict audio event sequences.

cs.SD↗

arXiv2022

Conformer-Based Self-Supervised Learning for Non-Speech Audio Tasks

Representation learning from unlabeled data has been of major interest in artificial intelligence research. While self-supervised speech representation learning has been popular in the speech research community, very few works have comprehensively analyzed audio representation learning for non-speech audio tasks. In this paper, we propose a self-supervised audio representation learning method and apply it to a variety of downstream non-speech audio tasks. We combine the well-known wav2vec 2.0 framework, which has shown success in self-supervised learning for speech tasks, with parameter-efficient conformer architectures. Our self-supervised pre-training can reduce the need for labeled data by two-thirds. On the AudioSet benchmark, we achieve a mean average precision (mAP) score of 0.415, which is a new state-of-the-art on this dataset through audio-only self-supervised learning. Our fine-tuned conformers also surpass or match the performance of previous systems pre-trained in a supervised way on several downstream tasks. We further discuss the important design considerations for both pre-training and fine-tuning.

cs.SD↗

arXiv2021

Illuminating Galaxy Evolution at Cosmic Noon with ISCEA: the Infrared Satellite for Cosmic Evolution Astrophysics

ISCEA (Infrared Satellite for Cosmic Evolution Astrophysics) is a small astrophysics mission whose Science Goal is to discover how galaxies evolved in the cosmic web of dark matter at cosmic noon. Its Science Objective is to determine the history of star formation and its quenching in galaxies as a function of local density and stellar mass when the Universe was 3-5 Gyrs old (1.2 10 Mpc in each of 50 protocluster (cluster and cosmic web) fields at 1.2 < z < 2.1. ISCEA measures the star formation quenching factor in those fields, and galaxy kinematics with a precision < 50 km/s to deduce the 3D spatial distribution in each field. ISCEA will transform our understanding of galaxy evolution at cosmic noon. ISCEA is a small satellite observatory with a 30cm equivalent diameter aperture telescope with a FoV of 0.32 deg^2, and a multi-object spectrograph with a digital micro-mirror device (DMD) as its programmable slit mask. ISCEA will obtain spectra of 1000 galaxies simultaneously at an effective resolving power of R=1000, with 2.8"x2.8" slits, over the NIR wavelength range of 1.1 to 2.0 microns, a regime not accessible from the ground without large gaps in coverage. ISCEA will achieve a pointing accuracy of <= 2" FWHM over 200s. ISCEA will be launched into a Low Earth Orbit, with a prime mission of 2.5 years. ISCEA's space-qualification of DMDs opens a new window for spectroscopy from space, enabling revolutionary advances in astrophysics.

astro-ph.GA↗

arXiv2021

Uniqueness and uniform structural stability of Poiseuille flows in an infinitely long pipe with Navier boundary conditions

In this paper, uniqueness and uniform structural stability of Poiseuille flows in an infinitely long pipe with Navier boundary conditions are established for axisymmetric solutions of steady Navier-Stokes system. The crucial point is that the estimate is uniform with respect both the flux of flows and slip coefficient which appeared in Navier boundary conditions. With the aid of special structure of Navier-Stokes system and the refined estimate for some quantities such as radial velocity, the uniqueness and existence of steady solutions of Navier-Stokes system can be obtained even when the external forces are large as long as the fluxes of flows are large. The delicate decomposition in the two dimensional plane for slip coefficient and frequency corresponding to Fourier variable in the axial direction plays a key role to achieve these estimates.

math.AP↗

arXiv2021

Uniqueness and uniform structural stability of Poiseuille flows in a periodic pipe with Navier boundary conditions

In this paper, we prove the uniqueness and structural stability of Poiseuille flows for axisymmetric solutions of steady Navier-Stokes system supplemented with Navier boundary conditions in a periodic pipe. Moreover, the stability is uniform with respect to both the flux and the slip coefficient of Navier boundary conditions. It is also showed that the non-zero frequency part of the velocity is bounded by a power function of the flux with negative power as long as the flux is suitably large. One of the key ingredients of the analysis is to prove the uniform linear structural stability, where the analysis for the boundary layers and the swirl velocity corresponding to the flux and the slip coefficients in different regimes plays a crucial role.

math.AP↗

arXiv2021

The spatial distribution deviation and the power suppression of baryons from dark matter

The spatial distribution between dark matter and baryonic matter of the Universe is biased or deviates from each other. In this work, by comparing the results derived from IllustrisTNG and WIGEON simulations, we find that many results obtained from TNG are similar to those from WIGEON data, but differences between the two simulations do exist. For the ratio of density power spectrum between dark matter and baryonic matter, as scales become smaller and smaller, the power spectra for baryons are increasingly suppressed for WIGEON simulations; while for TNG simulations, the suppression stops at $k=15-20h{\rm Mpc}^{-1}$, and the power spectrum ratios increase when $k>20h{\rm Mpc}^{-1}$. The suppression of power ratio for WIGEON is also redshift-dependent. From $z=1$ to $z=0$, the power ratio decreases from about 70% to less than 50% at $k=8h{\rm Mpc}^{-1}$. For TNG simulation, the suppression of power ratio is enhanced with decreasing redshifts in the scale range $k>4h{\rm Mpc}^{-1}$, but is nearly unchanged with redshifts in $k<4h{\rm Mpc}^{-1}$ These results indicate that turbulent heating can also have the consequence to suppress the power ratio between baryons and dark matter. Regarding the power suppression for TNG simulations as the norm, the power suppression by turbulence for WIGEON simulations is roughly estimated to be 45% at $k=2h{\rm Mpc}^{-1}$, and gradually increases to 69% at $k=8h{\rm Mpc}^{-1}$, indicating the impact of turbulence on the cosmic baryons are more significant on small scales.

astro-ph.CO↗

arXiv2021

Transferring Voice Knowledge for Acoustic Event Detection: An Empirical Study

Detection of common events and scenes from audio is useful for extracting and understanding human contexts in daily life. Prior studies have shown that leveraging knowledge from a relevant domain is beneficial for a target acoustic event detection (AED) process. Inspired by the observation that many human-centered acoustic events in daily life involve voice elements, this paper investigates the potential of transferring high-level voice representations extracted from a public speaker dataset to enrich an AED pipeline. Towards this end, we develop a dual-branch neural network architecture for the joint learning of voice and acoustic features during an AED process and conduct thorough empirical studies to examine the performance on the public AudioSet [1] with different types of inputs. Our main observations are that: 1) Joint learning of audio and voice inputs improves the AED performance (mean average precision) for both a CNN baseline (0.292 vs 0.134 mAP) and a TALNet [2] baseline (0.361 vs 0.351 mAP); 2) Augmenting the extra voice features is critical to maximize the model performance with dual inputs.

cs.SD↗

arXiv2021

The High Latitude Spectroscopic Survey on the Nancy Grace Roman Space Telescope

The Nancy Grace Roman Space Telescope will conduct a High Latitude Spectroscopic Survey (HLSS) over a large volume at high redshift, using the near-IR grism (1.0-1.93 $μ$m, $R=435-865$) and the 0.28 deg$^2$ wide field camera. We present a reference HLSS which maps 2000 deg$^2$ and achieves an emission line flux limit of 10$^{-16}$ erg/s/cm$^2$ at 6.5$σ$, requiring $\sim$0.6 yrs of observing time. We summarize the flowdown of the Roman science objectives to the science and technical requirements of the HLSS. We construct a mock redshift survey over the full HLSS volume by applying a semi-analytic galaxy formation model to a cosmological N-body simulation, and use this mock survey to create pixel-level simulations of 4 deg$^2$ of HLSS grism spectroscopy. We find that the reference HLSS would measure $\sim$ 10 million H$α$ galaxy redshifts that densely map large scale structure at $z=1-2$ and 2 million [OIII] galaxy redshifts that sparsely map structures at $z=2-3$. We forecast the performance of this survey for measurements of the cosmic expansion history with baryon acoustic oscillations and the growth of large scale structure with redshift space distortions. We also study possible deviations from the reference design, and find that a deep HLSS at $f_{\rm line}>7\times10^{-17}$erg/s/cm$^2$ over 4000 deg$^2$ (requiring $\sim$1.5 yrs of observing time) provides the most compelling stand-alone constraints on dark energy from Roman alone. This provides a useful reference for future optimizations. The reference survey, simulated data sets, and forecasts presented here will inform community decisions on the final scope and design of the Roman HLSS.

astro-ph.CO↗

arXiv2021

Simulating properties of emission line galaxies from Nancy Grace Roman Space Telescope

We simulate the emission line galaxy properties for the Nancy Grace Roman Space Telescope by creating a 4 deg$^{2}$ galaxy mock catalog using Galacticus, a semi-analytic model (SAM) for galaxy formation and evolution. The simulated galaxy properties include emission line luminosity, equivalent width (EW), broad band photometry, and spectral energy distribution (SED). We compare this mock catalog with the latest observational data from WISP and MOSDEF. We find that our Galacticus model makes predictions consistent with observational data, over a wide redshift range for multiple emission lines. We use this validated galaxy mock catalog to forecast the photometric completeness of H$α$ and [OIII] emission lines as a function of line flux cut and H band magnitdue cut, for both Roman and Euclid-like surveys. Our prediction for the photometric completeness of a Euclid-like survey is in good agreement with previous work based on WISP data. Our predictions for the photometric completeness of possible implementations of the Roman High Latitude Wide Area Spectroscopic Survey (HLWASS) provides key input for the planning of such a survey. We find that at H = 24, a Euclid-like survey to the line flux limit of 2$\times 10^{-16}\,$erg/s/cm$^2$ is 97\% complete, while a Roman HLWASS to the line flux limit of $10^{-16}\,$erg/s/cm$^2$ is only 94.6\% complete (it becomes 98\% complete at H = 25).

astro-ph.GA↗

arXiv2021

GRB 200716C: Evidence for a Short Burst Being Lensed

A tiny fraction of observed gamma-ray bursts (GRBs) may be lensed. The time delays induced by the gravitational lensing are milliseconds to seconds if the point lenses are intermediate-mass black holes. The prompt emission of the lensed GRBs, in principle, should have repeated pulses with identical light curves and spectra but different fluxes and slightly offset positions. In this work, we search for such candidates within the GRBs detected by Fermi/GBM, Swift/BAT, and HXMT/HE and report the identification of an attractive event GRB 200716C that consists of two pulses. Both the autocorrelation analysis and the Bayesian inference of the prompt emission light curve are in favor of the gravitational lensing scenario. Moreover, the spectral properties of the two pulses are rather similar and follow the so-called Amati relation of short GRBs rather than long duration bursts. The measured flux ratios between the two pulses are nearly constant in all channels, as expected from gravitational lensing. We therefore suggest that the long duration burst GRB 200716C was a short event being lensed. The redshifted mass of the lens was estimated to be $4.25^{+2.46}_{-1.36}$ $\times$ $10^5$ $M_{\odot}$ (90$\%$ credibility). If correct, this could point towards the existence of an intermediate-mass black hole along the line of sight of GRB 200716C.

astro-ph.HE↗

arXiv2021

The continuous wavelet derived by smoothing function and its application in cosmology

The wavelet analysis technique is a powerful tool and is widely used in broad disciplines of engineering, technology, and sciences. In this work, we present a novel scheme of constructing continuous wavelet functions, in which the wavelet functions are obtained by taking the first derivative of smoothing functions with respect to the scale parameter. Due to this wavelet constructing scheme, the inverse transforms are only one-dimensional integrations with respect to the scale parameter, and hence the continuous wavelet transforms constructed in this way are more ready to use than the usual scheme. We then apply the Gaussian-derived wavelet constructed by our scheme to computations of the density power spectrum for dark matter, the velocity power spectrum and the kinetic energy spectrum for baryonic fluid. These computations exhibit the convenience and strength of the continuous wavelet transforms. The transforms are very easy to perform, and we believe that the simplicity of our wavelet scheme will make continuous wavelet transforms very useful in practice.

astro-ph.IM↗

arXiv2021

AI4VIS: Survey on Artificial Intelligence Approaches for Data Visualization

Visualizations themselves have become a data format. Akin to other data formats such as text and images, visualizations are increasingly created, stored, shared, and (re-)used with artificial intelligence (AI) techniques. In this survey, we probe the underlying vision of formalizing visualizations as an emerging data format and review the recent advance in applying AI techniques to visualization data (AI4VIS). We define visualization data as the digital representations of visualizations in computers and focus on data visualization (e.g., charts and infographics). We build our survey upon a corpus spanning ten different fields in computer science with an eye toward identifying important common interests. Our resulting taxonomy is organized around WHAT is visualization data and its representation, WHY and HOW to apply AI to visualization data. We highlight a set of common tasks that researchers apply to the visualization data and present a detailed discussion of AI approaches developed to accomplish those tasks. Drawing upon our literature review, we discuss several important research questions surrounding the management and exploitation of visualization data, as well as the role of AI in support of those processes. We make the list of surveyed papers and related material available online at ai4vis.github.io.

cs.HC↗

arXiv2021

MultiVision: Designing Analytical Dashboards with Deep Learning Based Recommendation

We contribute a deep-learning-based method that assists in designing analytical dashboards for analyzing a data table. Given a data table, data workers usually need to experience a tedious and time-consuming process to select meaningful combinations of data columns for creating charts. This process is further complicated by the need of creating dashboards composed of multiple views that unveil different perspectives of data. Existing automated approaches for recommending multiple-view visualizations mainly build on manually crafted design rules, producing sub-optimal or irrelevant suggestions. To address this gap, we present a deep learning approach for selecting data columns and recommending multiple charts. More importantly, we integrate the deep learning models into a mixed-initiative system. Our model could make recommendations given optional user-input selections of data columns. The model, in turn, learns from provenance data of authoring logs in an offline manner. We compare our deep learning model with existing methods for visualization recommendation and conduct a user study to evaluate the usefulness of the system.

cs.HC↗

arXiv2021

Do sound event representations generalize to other audio tasks? A case study in audio transfer learning

Transfer learning is critical for efficient information transfer across multiple related learning problems. A simple, yet effective transfer learning approach utilizes deep neural networks trained on a large-scale task for feature extraction. Such representations are then used to learn related downstream tasks. In this paper, we investigate transfer learning capacity of audio representations obtained from neural networks trained on a large-scale sound event detection dataset. We build and evaluate these representations across a wide range of other audio tasks, via a simple linear classifier transfer mechanism. We show that such simple linear transfer is already powerful enough to achieve high performance on the downstream tasks. We also provide insights into the attributes of sound event representations that enable such efficient information transfer.

cs.SD↗

arXiv2021

PTNet: A High-Resolution Infant MRI Synthesizer Based on Transformer

Magnetic resonance imaging (MRI) noninvasively provides critical information about how human brain structures develop across stages of life. Developmental scientists are particularly interested in the first few years of neurodevelopment. Despite the success of MRI collection and analysis for adults, it is a challenge for researchers to collect high-quality multimodal MRIs from developing infants mainly because of their irregular sleep pattern, limited attention, inability to follow instructions to stay still, and a lack of analysis approaches. These challenges often lead to a significant reduction of usable data. To address this issue, researchers have explored various solutions to replace corrupted scans through synthesizing realistic MRIs. Among them, the convolution neural network (CNN) based generative adversarial network has demonstrated promising results and achieves state-of-the-art performance. However, adversarial training is unstable and may need careful tuning of regularization terms to stabilize the training. In this study, we introduced a novel MRI synthesis framework - Pyramid Transformer Net (PTNet). PTNet consists of transformer layers, skip-connections, and multi-scale pyramid representation. Compared with the most widely used CNN-based conditional GAN models (namely pix2pix and pix2pixHD), our model PTNet shows superior performance in terms of synthesis accuracy and model size. Notably, PTNet does not require any type of adversarial training and can be easily trained using the simple mean squared error loss.

eess.IV↗

arXiv2021

Linear bias and halo occupation distribution of emission line galaxies from Nancy Grace Roman Space Telescope

We present measurements of the linear galaxy bias of H$α$ and [OIII] emission line galaxies (ELGs) for the High Latitude Spectroscopic Survey (HLSS) of Nancy Grace Roman Space Telescope, using galaxy mocks constructed using semi-analytical model for galaxy formation, {\it Galacticus}, with a large cosmic volume and redshift coverage. We compute the two-point statistics of galaxies in configuration space and measure linear bias within scales of $10\sim50 h^{-1}$Mpc. We adopt different selection algorithms to investigate the impact of the Roman line flux cut, as well as the effect of dust model used to calibrate Galacticus, on the bias measurements. We consider galaxies with H$α$ and [OIII] emissions over the redshift range $1<z<3$, as specified by the current baseline for the Roman HLSS. We find that the linear bias for the H$α$ and [OIII] ELGs can be expressed as a linear function with respect to redshift: $b \simeq 0.88z+0.49$ for H$α$ $(1<z<2)$, and $b \simeq 0.98z+0.49$ for [OIII] $(2<z<3)$. We have also measured the Halo Occupation Distributions of these H$α$ and [OIII] emission line galaxies to understand their distribution within dark matter halos. Our results provide key input to enable the reliable forecast of dark energy and cosmology constraints from Roman.

astro-ph.CO↗

arXiv2021

Learning to Automate Chart Layout Configurations Using Crowdsourced Paired Comparison

We contribute a method to automate parameter configurations for chart layouts by learning from human preferences. Existing charting tools usually determine the layout parameters using predefined heuristics, producing sub-optimal layouts. People can repeatedly adjust multiple parameters (e.g., chart size, gap) to achieve visually appealing layouts. However, this trial-and-error process is unsystematic and time-consuming, without a guarantee of improvement. To address this issue, we develop Layout Quality Quantifier (LQ2), a machine learning model that learns to score chart layouts from pairwise crowdsourcing data. Combined with optimization techniques, LQ2 recommends layout parameters that improve the charts' layout quality. We apply LQ2 on bar charts and conduct user studies to evaluate its effectiveness by examining the quality of layouts it produces. Results show that LQ2 can generate more visually appealing layouts than both laypeople and baselines. This work demonstrates the feasibility and usages of quantifying human preferences and aesthetics for chart layouts.

cs.HC↗