arXiv subjects

Yun Zheng

Publications and source records attributed to Yun Zheng.

At least 55 records · Page 3Linked to original sources

arXiv2023

Photometric Objects Around Cosmic Webs (PAC) Delineated in a Spectroscopic Survey. IV. High Precision Constraints on the Evolution of Stellar-Halo Mass Relation at Redshift $z<0.7$

Taking advantage of the Photometric objects Around Cosmic webs (PAC) method developed in Paper I, we measure the excess surface density $\bar{n}_2w_{\rm{p}}$ of photometric objects around spectroscopic objects down to stellar mass $10^{8.0}M_{\odot}$, $10^{9.2}M_{\odot}$ and $10^{9.8}M_{\odot}$ in the redshift ranges of $z_s<0.2$, $0.2<z_s<0.4$ and $0.5<z_s<0.7$ respectively, using the data from the DESI Legacy Imaging Surveys and the spectroscopic samples of Slogan Digital Sky Survey (i.e. Main, LOWZ and CMASS samples). We model the measured $\bar{n}_2w_{\rm{p}}$ in N-body simulation using abundance matching method and constrain the stellar-halo mass relations (SHMR) in the three redshift ranges to percent level. With the accurate modeling, we demonstrate that the stellar mass scatter for given halo mass is nearly a constant, and that the empirical form of Behroozi et al describes the SHMR better than the double power law form at low mass. Our SHMR accurately captures the downsizing of massive galaxies since $z_s=0.7$, while it also indicates that small galaxies are still growing faster than their host halos. The galaxy stellar mass functions (GSMF) from our modeling are in perfect agreement with the {\it model-independent} measurements in Paper III, though the current work extends the GSMF to a much smaller stellar mass. Based on the GSMF and SHMR, we derive the stellar mass completeness and halo occupation distributions for the LOWZ and CMASS samples, which are useful for correctly interpreting their cosmological measurements such as galaxy-galaxy lensing and redshift space distortion.

astro-ph.GA↗

arXiv2022

eDIG-CHANGES I: Extended Hα Emission from the Extraplanar Diffuse Ionized Gas (eDIG) around CHANG-ES Galaxies

The extraplanar diffuse ionized gas (eDIG) represents the cool/warm ionized gas reservoir around galaxies. We present a spatial analysis of H$α$ images of 22 nearby edge-on spiral galaxies from the CHANG-ES sample (the eDIG-CHANGES project), taken with the APO 3.5m telescope, in order to study their eDIG. We conduct an exponential fit to the vertical intensity profiles of the sample galaxies, of which 16 can be decomposed into a thin disk plus an extended thick disk component. The median value of the scale height (h) of the extended component is $1.13\pm 0.14$ kpc. We find a tight sublinear correlation between h and the SFR. Moreover, the offset of individual galaxies from the best-fit SFR-h relation shows significant anti-correlation with SFR_SD. This indicates that galaxies with more intense star formation tend to have disproportionately extended eDIG. Combined with data from the literature, we find that the correlations between the eDIG properties and the galaxies' properties extend to broader ranges. We further compare the vertical extension of the eDIG to multi-wavelength measurements of other CGM phases. We find the eDIG to be slightly more extended than the neutral gas (HI 21-cm line), indicating the existence of some extended ionizing sources. Most galaxies have an X-ray scale height smaller than the h, suggesting that the majority of the X-ray emission detected in shallow observations are actually from the thick disk. The h is comparable to the L-band radio continuum scale height, both slightly larger than that at higher frequencies (C-band), where the cooling is stronger and the thermal contribution may be larger. The comparable H$α$ and L-band scale height indicates that the thermal and non-thermal electrons have similar spatial distributions. This further indicates that the thermal gas, the cosmics rays, and the magnetic field may be close to energy equipartition.

astro-ph.GA↗

arXiv2022

Gas dynamics and star formation in NGC 6822

We present H I gas kinematics and star formation activities of NGC 6822, a dwarf galaxy located in the Local Group at a distance of ~ 490 kpc. We perform profile decomposition of line-of-sight velocity profiles of the H I data cube (42.4" x 12.0" spatial, corresponding to ~ 100 pc; 1.6 km s$^{-1}$ spectral) taken with the Australia Telescope Compact Array (ATCA). For this, we use a new tool, the so-called BAYGAUD which is based on Bayesian analysis techniques, allowing us to decompose a line-of-sight velocity profile into an optimal number of Gaussian components in a quantitative manner. We classify the decomposed H I gas components of NGC 6822 into cool-bulk, warm-bulk, cool-non-bulk and warm-non-bulk motions with respect to their centroid velocities and velocity dispersions. We correlate their gas surface densities with corresponding star formation rate densities derived using both the GALEX far-ultraviolet and WISE 22 $μ$m data to examine the resolved Kennicutt-Schmidt (K-S) law for NGC 6822. Of the decomposed H I gas components, the cool-bulk component is likely to better follow the linear extension of the K-S law for molecular hydrogen (H$_2$) at low gas surface densities where H I is not saturated.

astro-ph.GA↗

arXiv2022

HI Vertical Structure of Nearby Edge-on Galaxies from CHANG-ES

We study the vertical distribution of the highly inclined galaxies from the Continuum Halos in Nearby Galaxies - an EVLA Survey (CHANG-ES). We explore the feasibility of photometrically deriving the HI disk scale-heights from the moment-0 images of the relatively edge-on galaxies with inclination >80 deg, by quantifying the systematic broadening effects and thus deriving correction equations for direct measurements. The corrected HI disk scale-heights of the relatively edge-on galaxies from the CHANG-ES sample show trends consistent with the quasi-equilibrium model of the vertical structure of gas disks. The procedure provide a convenient way to derive the scale-heights and can easily be applied to statistical samples in the future.

astro-ph.GA↗

arXiv2022

Observation of emergent Dirac physics at the surfaces of acoustic higher-order topological insulators

Using three-dimensional (3D) sonic crystals as acoustic higher-order topological insulators (HOTIs), we discover two-dimensional (2D) surface states described by spin-1 Dirac equations at the interfaces between the two sonic crystals with distinct topology but the same crystalline symmetry. We find that the Dirac mass can be tuned by the geometry of the two sonic crystals. The sign reversal of the Dirac mass reveals a surface topological transition where the surface states exhibit zero refractive index behavior. When the surface states are gapped, one-dimensional (1D) hinge states emerge due to the topology of the gapped surface states. We confirm experimentally the zero refractive index behavior and the emergent topological hinge states. Our study reveals a multidimensional Wannier orbital control that leads to extraordinary properties of surface states and unveils an interesting topological mechanism for the control of surface waves.

cond-mat.mes-hall↗

arXiv2022

CHANG-ES XXV: HI Imaging of Nearby Edge-on Galaxies -- Data Release 4

We present the HI distribution of galaxies from the Continuum Halos in Nearby Galaxies - an EVLA Survey (CHANG-ES). Though the observational mode was not optimized for detecting HI, we successfully produce HI cubes for 19 galaxies. The moment-0 maps from this work are available on CHANG-ES data release website, i.e., https://www.queensu.ca/changes. Our sample is dominated by star-forming, HI-rich galaxies at distances from 6.27 to 34.1 Mpc. HI interferometric images on two of these galaxies (NGC 5792 and UGC 10288) are presented here for the first time, while 12 of our remaining sample galaxies now have better HI spatial resolutions and/or sensitivities of intensity maps than those in existing publications. We characterize the average scale heights of the HI distributions for a subset of most inclined galaxies (inclination > 80 deg), and compare them to the radio continuum intensity scale heights, which have been derived in a similar way. The two types of scale heights are well correlated, with similar dependence on disk radial extension and star formation rate surface density but different dependence on mass surface density. This result indicates that the vertical distribution of the two components may be governed by similar fundamental physics but with subtle differences.

astro-ph.GA↗

arXiv2022

Disentangled Representation Learning for Text-Video Retrieval

Cross-modality interaction is a critical component in Text-Video Retrieval (TVR), yet there has been little examination of how different influencing factors for computing interaction affect performance. This paper first studies the interaction paradigm in depth, where we find that its computation can be split into two terms, the interaction contents at different granularity and the matching function to distinguish pairs with the same semantics. We also observe that the single-vector representation and implicit intensive function substantially hinder the optimization. Based on these findings, we propose a disentangled framework to capture a sequential and hierarchical representation. Firstly, considering the natural sequential structure in both text and video inputs, a Weighted Token-wise Interaction (WTI) module is performed to decouple the content and adaptively exploit the pair-wise correlations. This interaction can form a better disentangled manifold for sequential inputs. Secondly, we introduce a Channel DeCorrelation Regularization (CDCR) to minimize the redundancy between the components of the compared vectors, which facilitate learning a hierarchical representation. We demonstrate the effectiveness of the disentangled representation on various benchmarks, e.g., surpassing CLIP4Clip largely by +2.9%, +3.1%, +7.9%, +2.3%, +2.8% and +6.5% R@1 on the MSR-VTT, MSVD, VATEX, LSMDC, AcitivityNet, and DiDeMo, respectively.

cs.CV↗

arXiv2022

RCL: Recurrent Continuous Localization for Temporal Action Detection

Temporal representation is the cornerstone of modern action detection techniques. State-of-the-art methods mostly rely on a dense anchoring scheme, where anchors are sampled uniformly over the temporal domain with a discretized grid, and then regress the accurate boundaries. In this paper, we revisit this foundational stage and introduce Recurrent Continuous Localization (RCL), which learns a fully continuous anchoring representation. Specifically, the proposed representation builds upon an explicit model conditioned with video embeddings and temporal coordinates, which ensure the capability of detecting segments with arbitrary length. To optimize the continuous representation, we develop an effective scale-invariant sampling strategy and recurrently refine the prediction in subsequent iterations. Our continuous anchoring scheme is fully differentiable, allowing to be seamlessly integrated into existing detectors, e.g., BMN and G-TAD. Extensive experiments on two benchmarks demonstrate that our continuous representation steadily surpasses other discretized counterparts by ~2% mAP. As a result, RCL achieves 52.92% mAP@0.5 on THUMOS14 and 37.65% mAP on ActivtiyNet v1.3, outperforming all existing single-model detectors.

cs.CV↗

arXiv2022

Construct the emission line galaxy-host halo connection through auto and cross correlations

We investigate the [O\,II] emission line galaxy (ELG)-host halo connection via auto and cross correlations, and propose a concise and effective method to populate ELGs in dark matter halos without assuming a parameterized halo occupation distribution (HOD) model. Using the observational data from VIMOS Public Extragalactic Redshift Survey (VIPERS), we measure the auto and cross correlation functions between ELGs selected by [O\,II] luminosity and normal galaxies selected by stellar mass. Combining the stellar-halo mass relation (SHMR) derived for the normal galaxies and the fraction of ELGs observed in the normal galaxy population, we demonstrate that we can establish an accurate ELG-halo connection. With the ELG-halo connection, we can accurately reproduce the auto and cross correlation functions of ELGs and normal galaxies both in real-space and in redshift-space, once the satellite fraction is properly reduced. Our method provides a novel strategy to generate ELG mock catalogs for ongoing and upcoming galaxy redshift surveys. We also provide a simple description for the HOD of ELGs.

astro-ph.GA↗

arXiv2022

Photometric Objects around Cosmic Webs (PAC) Delineated in a Spectroscopic Survey. I. Methods

We provide a method for estimating the projected density distribution $\bar{n}_2w_p(r_p)$ of photometric objects around spectroscopic objects in a redshift survey. This quantity describes the distribution of Photometric sources with certain physical properties (e.g. luminosity, mass, color etc) Around Cosmic webs (PAC) traced by the spectroscopic objects. The method can make full use of current and future deep and wide photometric surveys to explore the formation of galaxies up to medium redshift ($z_s < 2$), with the aid of cosmological redshift surveys that sample only a fairly limited species of objects (e.g. Emission Line Galaxies). As an example, we apply the PAC method to the CMASS spectroscopic and HSC-SSP PDR2 photometric samples to explore the distribution of galaxies for a wide range of stellar mass from $10^{9.0}{\rm M_\odot}$ to $10^{12.0}{\rm M_\odot}$ around massive ones at $z_s\approx 0.6$. Using the abundance matching method, we model $\bar{n}_2w_p(r_p)$ in N-body simulation using MCMC sampling, and accurately measure the stellar-halo mass relation (SHMR) and stellar mass function (SMF) for the whole mass range. We can also measure the conditional stellar mass function (CSMF) of satellites for central galaxies of different mass. The PAC method has many potential applications for studying the evolution of galaxies.

astro-ph.GA↗

arXiv2022

Strong Conformity and Assembly Bias: Towards a Physical Understanding of the Galaxy-Halo Connection in SDSS Clusters

Understanding the physical connection between cluster galaxies and massive haloes is key to mitigating systematic uncertainties in next-generation cluster cosmology. We develop a novel method to infer the level of conformity between the stellar mass of the brightest central galaxies~(BCGs) $M_*^{BCG}$ and the satellite richness $λ$, defined as their correlation coefficient $ρ_{cc}$ at fixed halo mass, using the abundance and weak lensing of SDSS clusters as functions of $M_*^{BCG}$ and $λ$. We detect a halo mass-dependent conformity as $ρ_{cc}{=}0.60{+}0.08\ln(M_h/3{\times}10^{14}M_{\odot}/h)$. The strong conformity successfully resolves the "halo mass equality" conundrum discovered in Zu et al. 2021 -- when split by $M_*^{BCG}$ at fixed $λ$, the low and high-$M_*^{BCG}$ clusters have the same average halo mass despite having a $0.34$ dex discrepancy in average $M_*^{BCG}$. On top of the best-fitting conformity model, we develop a cluster assembly bias~(AB) prescription calibrated against the CosmicGrowth simulation, and build a conformity+AB model for the cluster weak lensing measurements. Our model predicts that with a ${\sim}20\%$ lower halo concentration $c$, the low-$M_*^{BCG}$ clusters are ${\sim}10\%$ more biased than the high-$M_*^{BCG}$ systems, in excellent agreement with the observations. We also show that the observed conformity and assembly bias are unlikely due to projection effects. Finally, we build a toy model to argue that while the early-time BCG-halo co-evolution drives the $M_*^{BCG}$-$c$ correlation, the late-time dry merger-induced BCG growth naturally produces the $M_*^{BCG}$-$λ$ conformity despite the well-known anti-correlation between $λ$ and $c$. Our method paves the path towards simultaneously constraining cosmology and cluster formation with future cluster surveys.

astro-ph.CO↗

arXiv2021

Measurements of $^{160}$Dy($p,γ$) at energies relevant for astrophysical $γ$ process

Rare information on photodisintegration reactions of nuclei with mass numbers $A \approx 160$ at astrophysical conditions impedes our understanding of the origin of $p$-nuclei. Experimental determination of the key ($p,γ$) cross sections has been playing an important role to verify nuclear reaction models and to provide rates of relevant ($γ,p$) reactions in $γ$-process. In this paper we report the first cross section measurements of $^{160}$Dy($p,γ$)$^{161}$Ho and $^{161}$Dy($p,n$)$^{161}$Ho in the beam energy range of 3.4 - 7.0 MeV, partially covering the Gamow window. Such determinations are possible by using two targets with various isotopic fractions. The cross section data can put a strong constraint on the nuclear level densities and gamma strength functions for $A \approx$ 160 in the Hauser-Feshbach statistical model. Furthermore, we find the best parameters for TALYS that reproduce the A $\thicksim$ 160 data available, $^{160}$Dy($p,γ$)$^{161}$Ho and $^{162}$Er($p,γ$)$^{163}$Tm, and recommend the constrained $^{161}$Ho($γ,p$)$^{160}$Dy reaction rates over a wide temperature range for $γ$-process network calculations. Although the determined $^{161}$Ho($γ$, p) stellar reaction rates at the temperature of 1 to 2 GK can differ by up to one order of magnitude from the NON-SMOKER predictions, it has a minor effect on the yields of $^{160}$Dy and accordingly the $p$-nuclei, $^{156,158}$Dy. A sensitivity study confirms that the cross section of $^{160}$Dy($p$, $γ$)$^{161}$Ho is measured precisely enough to predict yields of $p$-nuclei in the $γ$-process.

nucl-ex↗

arXiv2021

Multiple Object Tracking with Correlation Learning

Recent works have shown that convolutional networks have substantially improved the performance of multiple object tracking by simultaneously learning detection and appearance features. However, due to the local perception of the convolutional network structure itself, the long-range dependencies in both the spatial and temporal cannot be obtained efficiently. To incorporate the spatial layout, we propose to exploit the local correlation module to model the topological relationship between targets and their surrounding environment, which can enhance the discriminative power of our model in crowded scenes. Specifically, we establish dense correspondences of each spatial location and its context, and explicitly constrain the correlation volumes through self-supervised learning. To exploit the temporal context, existing approaches generally utilize two or more adjacent frames to construct an enhanced feature representation, but the dynamic motion scene is inherently difficult to depict via CNNs. Instead, our paper proposes a learnable correlation operator to establish frame-to-frame matches over convolutional feature maps in the different layers to align and propagate temporal context. With extensive experimental results on the MOT datasets, our approach demonstrates the effectiveness of correlation learning with the superior performance and obtains state-of-the-art MOTA of 76.5% and IDF1 of 73.6% on MOT17.

cs.CV↗

arXiv2021

Few-Shot Incremental Learning with Continually Evolved Classifiers

Few-shot class-incremental learning (FSCIL) aims to design machine learning algorithms that can continually learn new concepts from a few data points, without forgetting knowledge of old classes. The difficulty lies in that limited data from new classes not only lead to significant overfitting issues but also exacerbate the notorious catastrophic forgetting problems. Moreover, as training data come in sequence in FSCIL, the learned classifier can only provide discriminative information in individual sessions, while FSCIL requires all classes to be involved for evaluation. In this paper, we address the FSCIL problem from two aspects. First, we adopt a simple but effective decoupled learning strategy of representations and classifiers that only the classifiers are updated in each incremental session, which avoids knowledge forgetting in the representations. By doing so, we demonstrate that a pre-trained backbone plus a non-parametric class mean classifier can beat state-of-the-art methods. Second, to make the classifiers learned on individual sessions applicable to all classes, we propose a Continually Evolved Classifier (CEC) that employs a graph model to propagate context information between classifiers for adaptation. To enable the learning of CEC, we design a pseudo incremental learning paradigm that episodically constructs a pseudo incremental learning task to optimize the graph parameters by sampling data from the base dataset. Experiments on three popular benchmark datasets, including CIFAR100, miniImageNet, and Caltech-USCD Birds-200-2011 (CUB200), show that our method significantly outperforms the baselines and sets new state-of-the-art results with remarkable advantages.

cs.CV↗

arXiv2021

Fashion Focus: Multi-modal Retrieval System for Video Commodity Localization in E-commerce

Nowadays, live-stream and short video shopping in E-commerce have grown exponentially. However, the sellers are required to manually match images of the selling products to the timestamp of exhibition in the untrimmed video, resulting in a complicated process. To solve the problem, we present an innovative demonstration of multi-modal retrieval system called "Fashion Focus", which enables to exactly localize the product images in the online video as the focuses. Different modality contributes to the community localization, including visual content, linguistic features and interaction context are jointly investigated via presented multi-modal learning. Our system employs two procedures for analysis, including video content structuring and multi-modal retrieval, to automatically achieve accurate video-to-shop matching. Fashion Focus presents a unified framework that can orientate the consumers towards relevant product exhibitions during watching videos and help the sellers to effectively deliver the products over search and recommendation.

cs.CV↗

arXiv2021

Visual Search at Alibaba

This paper introduces the large scale visual search algorithm and system infrastructure at Alibaba. The following challenges are discussed under the E-commercial circumstance at Alibaba (a) how to handle heterogeneous image data and bridge the gap between real-shot images from user query and the online images. (b) how to deal with large scale indexing for massive updating data. (c) how to train deep models for effective feature representation without huge human annotations. (d) how to improve the user engagement by considering the quality of the content. We take advantage of large image collection of Alibaba and state-of-the-art deep learning techniques to perform visual search at scale. We present solutions and implementation details to overcome those problems and also share our learnings from building such a large scale commercial visual search engine. Specifically, model and search-based fusion approach is introduced to effectively predict categories. Also, we propose a deep CNN model for joint detection and feature learning by mining user click behavior. The binary index engine is designed to scale up indexing without compromising recall and precision. Finally, we apply all the stages into an end-to-end system architecture, which can simultaneously achieve highly efficient and scalable performance adapting to real-shot images. Extensive experiments demonstrate the advancement of each module in our system. We hope visual search at Alibaba becomes more widely incorporated into today's commercial applications.

cs.CV↗

arXiv2021

Large Scale Long-tailed Product Recognition System at Alibaba

A practical large scale product recognition system suffers from the phenomenon of long-tailed imbalanced training data under the E-commercial circumstance at Alibaba. Besides product images at Alibaba, plenty of image related side information (e.g. title, tags) reveal rich semantic information about images. Prior works mainly focus on addressing the long tail problem in visual perspective only, but lack of consideration of leveraging the side information. In this paper, we present a novel side information based large scale visual recognition co-training~(SICoT) system to deal with the long tail problem by leveraging the image related side information. In the proposed co-training system, we firstly introduce a bilinear word attention module aiming to construct a semantic embedding over the noisy side information. A visual feature and semantic embedding co-training scheme is then designed to transfer knowledge from classes with abundant training data (head classes) to classes with few training data (tail classes) in an end-to-end fashion. Extensive experiments on four challenging large scale datasets, whose numbers of classes range from one thousand to one million, demonstrate the scalable effectiveness of the proposed SICoT system in alleviating the long tail problem. In the visual search platform Pailitao\footnote{http://www.pailitao.com} at Alibaba, we settle a practical large scale product recognition application driven by the proposed SICoT system, and achieve a significant gain of unique visitor~(UV) conversion rate.

cs.CV↗

arXiv2021

Virtual ID Discovery from E-commerce Media at Alibaba: Exploiting Richness of User Click Behavior for Visual Search Relevance

Visual search plays an essential role for E-commerce. To meet the search demands of users and promote shopping experience at Alibaba, visual search relevance of real-shot images is becoming the bottleneck. Traditional visual search paradigm is usually based upon supervised learning with labeled data. However, large-scale categorical labels are required with expensive human annotations, which limits its applicability and also usually fails in distinguishing the real-shot images. In this paper, we propose to discover Virtual ID from user click behavior to improve visual search relevance at Alibaba. As a totally click-data driven approach, we collect various types of click data for training deep networks without any human annotations at all. In particular, Virtual ID are learned as classification supervision with co-click embedding, which explores image relationship from user co-click behaviors to guide category prediction and feature learning. Concretely, we deploy Virtual ID Category Network by integrating first-clicks and switch-clicks as regularizer. Incorporating triplets and list constraints, Virtual ID Feature Network is trained in a joint classification and ranking manner. Benefiting from exploration of user click data, our networks are more effective to encode richer supervision and better distinguish real-shot images in terms of category and feature. To validate our method for visual search relevance, we conduct an extensive set of offline and online experiments on the collected real-shot images. We consistently achieve better experimental results across all components, compared with alternative and state-of-the-art methods.

cs.CV↗