arXiv subjects

Cheng Li

Publications and source records attributed to Cheng Li.

At least 397 records · Page 22Linked to original sources

arXiv2019

Extensive beam test study of prototype MRPCs for the T0 detector at the CSR external-target experiment

The CSR External-target Experiment (CEE) will be the first large-scale nuclear physics experiment device at the Cooling Storage Ring (CSR) of the Heavy-Ion Research Facility in Lanzhou (HIRFL) in China. A new T0 detector has been proposed to measure the multiplicity, angular distribution and timing information of charged particles produced in heavy-ion collisions at the target region. Multi-gap resistive plate chamber (MRPC) technology was chosen as part of the construction of the T0 detector, which provides precision event collision times (T0) and collision geometry information. The prototype was tested with hadron and heavy-ion beams to study its performance. By comparing the experimental results with a Monte Carlo simulation, the time resolution of the MRPCs are found to be $\sim$ 50 ps or better. The timing performance of the T0 detector, including both detector and readout electronics, we found to fulfil the requirements of the CEE.

physics.ins-det↗

arXiv2019

The ALMaQUEST Survey: The molecular gas main sequence and the origin of the star forming main sequence

The origin of the star forming main sequence ( i.e., the relation between star formation rate and stellar mass, globally or on kpc-scales; hereafter SFMS) remains a hotly debated topic in galaxy evolution. Using the ALMA-MaNGA QUEnching and STar formation (ALMaQUEST) survey, we show that for star forming spaxels in the main sequence galaxies, the three local quantities, star-formation rate surface density (\sigsfr), stellar mass surface density (\sigsm), and the \h2~mass surface density (\sigh2), are strongly correlated with one another and form a 3D linear (in log) relation with dispersion. In addition to the two well known scaling relations, the resolved SFMS (\sigsfr~ vs. \sigsm) and the Schmidt-Kennicutt relation (\sigsfr~ vs. \sigh2; SK relation), there is a third scaling relation between \sigh2~ and \sigsm, which we refer to as the `molecular gas main sequence' (MGMS). The latter indicates that either the local gas mass traces the gravitational potential set by the local stellar mass or both quantities follow the underlying total mass distributions. The scatter of the resolved SFMS ($σ\sim 0.25$ dex) is the largest compared to those of the SK and MGMS relations ($σ\sim$ 0.2 dex). A Pearson correlation test also indicates that the SK and MGMS relations are more strongly correlated than the resolved SFMS. Our result suggests a scenario in which the resolved SFMS is the least physically fundamental and is the consequence of the combination of the SK and the MGMS relations.

astro-ph.GA↗

arXiv2019

AI Matrix: A Deep Learning Benchmark for Alibaba Data Centers

Alibaba has China's largest e-commerce platform. To support its diverse businesses, Alibaba has its own large-scale data centers providing the computing foundation for a wide variety of software applications. Among these applications, deep learning (DL) has been playing an important role in delivering services like image recognition, objection detection, text recognition, recommendation, and language processing. To build more efficient data centers that deliver higher performance for these DL applications, it is important to understand their computational needs and use that information to guide the design of future computing infrastructure. An effective way to achieve this is through benchmarks that can fully represent Alibaba's DL applications.

cs.LG↗

arXiv2019

The CFHT Large Area U-band Deep Survey (CLAUDS)

The CFHT Large Area U-band Deep Survey (CLAUDS) uses data taken with the MegaCam mosaic imager on CFHT to produce images of 18.60 deg2 with median seeing of FWHM=0.92 arcsec and to a median depth of U = 27.1 AB (5 sigma in 2 arcsec apertures), with selected areas that total 1.36 deg2 reaching a median depth of U=27.7 AB. These are the deepest U-band images assembled to date over this large an area. These data are located in four fields also imaged to comparably faint levels in grizy and several narrowband filters as part of the Hyper Suprime-Cam (HSC) Subaru Strategic Program (HSC-SSP). These CFHT and Subaru datasets will remain unmatched in their combination of area and depth until the advent of the Large Synoptic Survey Telescope (LSST). This paper provides an overview of the scientific motivation for CLAUDS and gives details of the observing strategy, observations, data reduction, and data merging with the HSC-SSP. Three early applications of these deep data are used to illustrate the potential of the dataset: deep U-band galaxy number counts, z~3 Lyman break galaxy (LBG) selection, and photometric redshifts improved by adding CLAUDS U to the Subaru HSC grizy photometry.

astro-ph.GA↗

arXiv2019

SDSS-IV MaNGA: Environmental dependence of gas metallicity gradients in local star-forming galaxies

Within the standard model of hierarchical galaxy formation in a ΛCDM Universe, the environment of galaxies is expected to play a key role in driving galaxy formation and evolution. In this paper we investigate whether and how the gas metallicity and the star formation surface density (Σ_SFR) depend on galaxy environment. To this end we analyse a sample of 1162 local, star-forming galaxies from the galaxy survey Mapping Nearby Galaxies at APO (MaNGA). Generally, both parameters do not show any significant dependence on environment. However, in agreement with previous studies, we find that low-mass satellite galaxies are an exception to this rule. The gas metallicity in these objects increases while their ΣSFR decreases slightly with environmental density. The present analysis of MaNGA data allows us to extend this to spatially resolved properties. Our study reveals that the gas metallicity gradients of low-mass satellites flatten and their ΣSFR gradients steepen with increasing environmental density. By extensively exploring a chemical evolution model, we identify two scenarios that are able to explain this pattern: metal-enriched gas accretion or pristine gas inflow with varying accretion timescales. The latter scenario better matches the observed ΣSFR gradients, and is therefore our preferred solution. In this model, a shorter gas accretion timescale at larger radii is required. This suggests that 'outside-in quenching' governs the star formation processes of low-mass satellite galaxies in dense environments.

astro-ph.GA↗

arXiv2019

Delving Deep Into Hybrid Annotations for 3D Human Recovery in the Wild

Though much progress has been achieved in single-image 3D human recovery, estimating 3D model for in-the-wild images remains a formidable challenge. The reason lies in the fact that obtaining high-quality 3D annotations for in-the-wild images is an extremely hard task that consumes enormous amount of resources and manpower. To tackle this problem, previous methods adopt a hybrid training strategy that exploits multiple heterogeneous types of annotations including 3D and 2D while leaving the efficacy of each annotation not thoroughly investigated. In this work, we aim to perform a comprehensive study on cost and effectiveness trade-off between different annotations. Specifically, we focus on the challenging task of in-the-wild 3D human recovery from single images when paired 3D annotations are not fully available. Through extensive experiments, we obtain several observations: 1) 3D annotations are efficient, whereas traditional 2D annotations such as 2D keypoints and body part segmentation are less competent in guiding 3D human recovery. 2) Dense Correspondence such as DensePose is effective. When there are no paired in-the-wild 3D annotations available, the model exploiting dense correspondence can achieve 92% of the performance compared to a model trained with paired 3D data. We show that incorporating dense correspondence into in-the-wild 3D human recovery is promising and competitive due to its high efficiency and relatively low annotating cost. Our model trained with dense correspondence can serve as a strong reference for future research.

cs.CV↗

arXiv2019

Deep Comprehensive Correlation Mining for Image Clustering

Recent developed deep unsupervised methods allow us to jointly learn representation and cluster unlabelled data. These deep clustering methods mainly focus on the correlation among samples, e.g., selecting high precision pairs to gradually tune the feature representation, which neglects other useful correlations. In this paper, we propose a novel clustering framework, named deep comprehensive correlation mining(DCCM), for exploring and taking full advantage of various kinds of correlations behind the unlabeled data from three aspects: 1) Instead of only using pair-wise information, pseudo-label supervision is proposed to investigate category information and learn discriminative features. 2) The features' robustness to image transformation of input space is fully explored, which benefits the network learning and significantly improves the performance. 3) The triplet mutual information among features is presented for clustering problem to lift the recently discovered instance-level deep mutual information to a triplet-level formation, which further helps to learn more discriminative features. Extensive experiments on several challenging datasets show that our method achieves good performance, e.g., attaining $62.3\%$ clustering accuracy on CIFAR-10, which is $10.1\%$ higher than the state-of-the-art results.

cs.CV↗

arXiv2019

Earth as an Exoplanet: A Two-dimensional Alien Map

Resolving spatially-varying exoplanet features from single-point light curves is essential for determining whether Earth-like worlds harbor geological features and/or climate systems that influence habitability. To evaluate the feasibility and requirements of this spatial feature resolving problem, we present an analysis of multi-wavelength single-point light curves of Earth, where it plays the role of a proxy exoplanet. Here, ~10,000 DSCOVR/EPIC frames collected over a two-year period were integrated over the Earth's disk to yield a spectrally-dependent point source and analyzed using singular value decomposition. We found that, between the two dominant principal components (PCs), the second PC contains surface-related features of the planet, while the first PC mainly includes cloud information. We present the first two-dimensional (2D) surface map of Earth reconstructed from light curve observations without any assumptions of its spectral properties. This study serves as a baseline for reconstructing the surface features of Earth-like exoplanets in the future.

astro-ph.EP↗

arXiv2019

AUNet: Attention-guided dense-upsampling networks for breast mass segmentation in whole mammograms

Mammography is one of the most commonly applied tools for early breast cancer screening. Automatic segmentation of breast masses in mammograms is essential but challenging due to the low signal-to-noise ratio and the wide variety of mass shapes and sizes. Existing methods deal with these challenges mainly by extracting mass-centered image patches manually or automatically. However, manual patch extraction is time-consuming and automatic patch extraction brings errors that could not be compensated in the following segmentation step. In this study, we propose a novel attention-guided dense-upsampling network (AUNet) for accurate breast mass segmentation in whole mammograms directly. In AUNet, we employ an asymmetrical encoder-decoder structure and propose an effective upsampling block, attention-guided dense-upsampling block (AU block). Especially, the AU block is designed to have three merits. Firstly, it compensates the information loss of bilinear upsampling by dense upsampling. Secondly, it designs a more effective method to fuse high- and low-level features. Thirdly, it includes a channel-attention function to highlight rich-information channels. We evaluated the proposed method on two publicly available datasets, CBIS-DDSM and INbreast. Compared to three state-of-the-art fully convolutional networks, AUNet achieved the best performances with an average Dice similarity coefficient of 81.8% for CBIS-DDSM and 79.1% for INbreast.

cs.CV↗

arXiv2019

Learning Cross-Modal Deep Representations for Multi-Modal MR Image Segmentation

Multi-modal magnetic resonance imaging (MRI) is essential in clinics for comprehensive diagnosis and surgical planning. Nevertheless, the segmentation of multi-modal MR images tends to be time-consuming and challenging. Convolutional neural network (CNN)-based multi-modal MR image analysis commonly proceeds with multiple down-sampling streams fused at one or several layers. Although inspiring performance has been achieved, the feature fusion is usually conducted through simple summation or concatenation without optimization. In this work, we propose a supervised image fusion method to selectively fuse the useful information from different modalities and suppress the respective noise signals. Specifically, an attention block is introduced as guidance for the information selection. From the different modalities, one modality that contributes most to the results is selected as the master modality, which supervises the information selection of the other assistant modalities. The effectiveness of the proposed method is confirmed through breast mass segmentation in MR images of two modalities and better segmentation results are achieved compared to the state-of-the-art methods.

eess.IV↗

arXiv2019

Model-based Convolutional De-Aliasing Network Learning for Parallel MR Imaging

Parallel imaging has been an essential technique to accelerate MR imaging. Nevertheless, the acceleration rate is still limited due to the ill-condition and challenges associated with the undersampled reconstruction. In this paper, we propose a model-based convolutional de-aliasing network with adaptive parameter learning to achieve accurate reconstruction from multi-coil undersampled k-space data. Three main contributions have been made: a de-aliasing reconstruction model was proposed to accelerate parallel MR imaging with deep learning exploring both spatial redundancy and multi-coil correlations; a split Bregman iteration algorithm was developed to solve the model efficiently; and unlike most existing parallel imaging methods which rely on the accuracy of the estimated multi-coil sensitivity, the proposed method can perform parallel reconstruction from undersampled data without explicit sensitivity calculation. Evaluations were conducted on \emph{in vivo} brain dataset with a variety of undersampling patterns and different acceleration factors. Our results demonstrated that this method could achieve superior performance in both quantitative and qualitative analysis, compared to three state-of-the-art methods.

eess.IV↗

arXiv2019

One-shot Face Reenactment

To enable realistic shape (e.g. pose and expression) transfer, existing face reenactment methods rely on a set of target faces for learning subject-specific traits. However, in real-world scenario end-users often only have one target face at hand, rendering existing methods inapplicable. In this work, we bridge this gap by proposing a novel one-shot face reenactment learning framework. Our key insight is that the one-shot learner should be able to disentangle and compose appearance and shape information for effective modeling. Specifically, the target face appearance and the source face shape are first projected into latent spaces with their corresponding encoders. Then these two latent spaces are associated by learning a shared decoder that aggregates multi-level features to produce the final reenactment results. To further improve the synthesizing quality on mustache and hair regions, we additionally propose FusionNet which combines the strengths of our learned decoder and the traditional warping method. Extensive experiments show that our one-shot face reenactment system achieves superior transfer fidelity as well as identity preserving capability than alternatives. More remarkably, our approach trained with only one target image per subject achieves competitive results to those using a set of target images, demonstrating the practical merit of this work. Code, models and an additional set of reenacted faces have been publicly released at the project page.

cs.CV↗

arXiv2019

Investigation on the Atmospheric Incoming Flow of a Utility-Scale Wind Turbine using Super-large-scale Particle Image Velocimetry

The atmospheric incoming flow of a wind turbine is intimately connected to its power production as well as its structural stability. Here we present an incoming flow measurement of a utility-scale turbine at the high spatio-temporal resolution, using super-large-scale particle image velocimetry (SLPIV) with natural snowflakes. The datasets include over a one-hour duration of incoming flow with a field of view of 85 m (vertical) x 40 m (streamwise) centered at 0.2 rotor diameter upstream of the turbine. The mean flow shows the presence of the induction zone and a distinct region with enhanced vertical velocity. Time series of nacelle sonic anemometer and SLPIV measured streamwise velocity outside the induction zone show generally matched trends with time-varying discrepancies potentially due to the induction effect and the flow acceleration around the nacelle. These discrepancies between the two signals, characterized by the sonic-SLPIV velocity ratio, is normally distributed and is less than unity 85% of the time. The velocity ratio first decreases with increasing wind speed up to around the rated speed of the turbine, then plateaus, and finally rises with a further increase in wind speed. With conditional sampling, the distribution of the velocity ratio shows that larger yaw error leads to an increase in both the mean and the spread of the distribution. Moreover, as the incident angle of the incoming flow changes from negative to positive (i.e. from pointing downward to upward), the velocity ratio first decreases as the angle approaches zero. With further increase of the incidence angle, the ratio then plateaus and fluctuations are augmented. Finally, our results show that the intensity of short-term velocity fluctuation has a limited impact on the sonic-SLPIV velocity ratio.

physics.flu-dyn↗

arXiv2019

Data-driven subtraction of anisotropic flows in jet-like correlation studies in heavy-ion collisions

Measurements of two-particle azimuthal angle correlations are a useful tool to study the distribution of jet energy loss, however, they are complicated because of the significant anisotropic flow background. We devise a data-driven method for subtracting anisotropic flow background in jet-like correlation analysis. We first require a large recoil momentum ($P_x$) within a given pseudo-rapidity ($η$) range from a high-transverse momentum particle to enhance in-acceptance population of away-side jet-like correlations. Then we take the difference of two-particle correlations in the close-region and far-region with respect to the $η$ region of $P_x$ to subtract the anisotropic flow background. We use a toy model which contains only anisotropic flow and PYTHIA8 which have jets to demonstrate the validity of our data-driven method. The results indicate that the data-driven method can subtract anisotropic flow effectively.

nucl-ex↗

arXiv2019

Accelerating Experimental Design by Incorporating Experimenter Hunches

Experimental design is a process of obtaining a product with target property via experimentation. Bayesian optimization offers a sample-efficient tool for experimental design when experiments are expensive. Often, expert experimenters have 'hunches' about the behavior of the experimental system, offering potentials to further improve the efficiency. In this paper, we consider per-variable monotonic trend in the underlying property that results in a unimodal trend in those variables for a target value optimization. For example, sweetness of a candy is monotonic to the sugar content. However, to obtain a target sweetness, the utility of the sugar content becomes a unimodal function, which peaks at the value giving the target sweetness and falls off both ways. In this paper, we propose a novel method to solve such problems that achieves two main objectives: a) the monotonicity information is used to the fullest extent possible, whilst ensuring that b) the convergence guarantee remains intact. This is achieved by a two-stage Gaussian process modeling, where the first stage uses the monotonicity trend to model the underlying property, and the second stage uses `virtual' samples, sampled from the first, to model the target value optimization function. The process is made theoretically consistent by adding appropriate adjustment factor in the posterior computation, necessitated because of using the `virtual' samples. The proposed method is evaluated through both simulations and real world experimental design problems of a) new short polymer fiber with the target length, and b) designing of a new three dimensional porous scaffolding with a target porosity. In all scenarios our method demonstrates faster convergence than the basic Bayesian optimization approach not using such `hunches'.

stat.ML↗

arXiv2019

CLCI-Net: Cross-Level fusion and Context Inference Networks for Lesion Segmentation of Chronic Stroke

Segmenting stroke lesions from T1-weighted MR images is of great value for large-scale stroke rehabilitation neuroimaging analyses. Nevertheless, there are great challenges with this task, such as large range of stroke lesion scales and the tissue intensity similarity. The famous encoder-decoder convolutional neural network, which although has made great achievements in medical image segmentation areas, may fail to address these challenges due to the insufficient uses of multi-scale features and context information. To address these challenges, this paper proposes a Cross-Level fusion and Context Inference Network (CLCI-Net) for the chronic stroke lesion segmentation from T1-weighted MR images. Specifically, a Cross-Level feature Fusion (CLF) strategy was developed to make full use of different scale features across different levels; Extending Atrous Spatial Pyramid Pooling (ASPP) with CLF, we have enriched multi-scale features to handle the different lesion sizes; In addition, convolutional long short-term memory (ConvLSTM) is employed to infer context information and thus capture fine structures to address the intensity similarity issue. The proposed approach was evaluated on an open-source dataset, the Anatomical Tracings of Lesions After Stroke (ATLAS) with the results showing that our network outperforms five state-of-the-art methods. We make our code and models available at https://github.com/YH0517/CLCI_Net.

eess.IV↗

arXiv2019

Frustrated with Replicating Claims of a Shared Model? A Solution

Machine Learning (ML) and Deep Learning (DL) innovations are being introduced at such a rapid pace that model owners and evaluators are hard-pressed analyzing and studying them. This is exacerbated by the complicated procedures for evaluation. The lack of standard systems and efficient techniques for specifying and provisioning ML/DL evaluation is the main cause of this "pain point". This work discusses common pitfalls for replicating DL model evaluation, and shows that these subtle pitfalls can affect both accuracy and performance. It then proposes a solution to remedy these pitfalls called MLModelScope, a specification for repeatable model evaluation and a runtime to provision and measure experiments. We show that by easing the model specification and evaluation process, MLModelScope facilitates rapid adoption of ML/DL innovations.

cs.LG↗

arXiv2019

Challenges and Pitfalls of Machine Learning Evaluation and Benchmarking

An increasingly complex and diverse collection of Machine Learning (ML) models as well as hardware/software stacks, collectively referred to as "ML artifacts", are being proposed - leading to a diverse landscape of ML. These ML innovations proposed have outpaced researchers' ability to analyze, study and adapt them. This is exacerbated by the complicated and sometimes non-reproducible procedures for ML evaluation. A common practice of sharing ML artifacts is through repositories where artifact authors post ad-hoc code and some documentation, but often fail to reveal critical information for others to reproduce their results. This results in users' inability to compare with artifact authors' claims or adapt the model to his/her own use. This paper discusses common challenges and pitfalls of ML evaluation and benchmarking, which can be used as a guideline for ML model authors when sharing ML artifacts, and for system developers when benchmarking or designing ML systems.

cs.LG↗