arXiv subjects

Yan Xu

Publications and source records attributed to Yan Xu.

At least 181 records · Page 10Linked to original sources

arXiv2023

Towards Answering Open-ended Ethical Quandary Questions

Considerable advancements have been made in various NLP tasks based on the impressive power of large language models (LLMs) and many NLP applications are deployed in our daily lives. In this work, we challenge the capability of LLMs with the new task of Ethical Quandary Generative Question Answering. Ethical quandary questions are more challenging to address because multiple conflicting answers may exist to a single quandary. We explore the current capability of LLMs in providing an answer with a deliberative exchange of different perspectives to an ethical quandary, in the approach of Socratic philosophy, instead of providing a closed answer like an oracle. We propose a model that searches for different ethical principles applicable to the ethical quandary and generates an answer conditioned on the chosen principles through prompt-based few-shot learning. We also discuss the remaining challenges and ethical issues involved in this task and suggest the direction toward developing responsible NLP systems by incorporating human values explicitly.

cs.CL↗

arXiv2022

PDNet: Toward Better One-Stage Object Detection With Prediction Decoupling

Recent one-stage object detectors follow a per-pixel prediction approach that predicts both the object category scores and boundary positions from every single grid location. However, the most suitable positions for inferring different targets, i.e., the object category and boundaries, are generally different. Predicting all these targets from the same grid location thus may lead to sub-optimal results. In this paper, we analyze the suitable inference positions for object category and boundaries, and propose a prediction-target-decoupled detector named PDNet to establish a more flexible detection paradigm. Our PDNet with the prediction decoupling mechanism encodes different targets separately in different locations. A learnable prediction collection module is devised with two sets of dynamic points, i.e., dynamic boundary points and semantic points, to collect and aggregate the predictions from the favorable regions for localization and classification. We adopt a two-step strategy to learn these dynamic point positions, where the prior positions are estimated for different targets first, and the network further predicts residual offsets to the positions with better perceptions of the object properties. Extensive experiments on the MS COCO benchmark demonstrate the effectiveness and efficiency of our method. With a single ResNeXt-64x4d-101-DCN as the backbone, our detector achieves 50.1 AP with single-scale testing, which outperforms the state-of-the-art methods by an appreciable margin under the same experimental settings.Moreover, our detector is highly efficient as a one-stage framework. Our code is public at https://github.com/yangli18/PDNet.

cs.CV↗

arXiv2022

Solar Flare Ribbon Fronts I: Constraining flare energy deposition with IRIS spectroscopy

Lower atmospheric lines show peculiar profiles at the leading edge of ribbons during solar flares. In particular, increased absorption of the BBSO/GST \hei~10830~Å line \citep[e.g.][]{Xu2016}, as well as broad and centrally reversed profiles in the spectra of the \mgii~and \cii~lines observed by the \iris~satellite \citep[e.g.][]{Panos2018,Panos2021a} have been reported. In this work, we aim to understand the physical origin of the \iris\ ribbon front line profiles, which seem to be common of many, if not all, flares. To achieve this, we quantify the spectral properties of the \iris~\mgii~ribbon front profiles during four large flares and perform a detailed comparison with a grid of radiative hydrodynamic models using the \radynfp~code. We also studied their transition region counterparts, finding that these ribbon front locations are regions where transition region emission and chromospheric evaporation are considerably weaker compared to other parts of the ribbons. Based on our comparison between the \iris~observations and modelling, our interpretation is that there are different heating regimes at play in the leading and trailing regions of the ribbons. More specifically, we suggest that bombardment of the chromosphere by more gradual and modest non-thermal electron energy fluxes can qualitatively explain the \iris~observations at the ribbon front, while stronger and more impulsive energy fluxes are required to drive chromospheric evaporation and more intense TR emission. Our results provide a possible physical origin for the peculiar behaviour of the \iris~chromospheric lines in the ribbon leading edge and new constraints for the flare models.

astro-ph.SR↗

arXiv2022

A Deep Learning Approach to Generating Photospheric Vector Magnetograms of Solar Active Regions for SOHO/MDI Using SDO/HMI and BBSO Data

Solar activity is usually caused by the evolution of solar magnetic fields. Magnetic field parameters derived from photospheric vector magnetograms of solar active regions have been used to analyze and forecast eruptive events such as solar flares and coronal mass ejections. Unfortunately, the most recent solar cycle 24 was relatively weak with few large flares, though it is the only solar cycle in which consistent time-sequence vector magnetograms have been available through the Helioseismic and Magnetic Imager (HMI) on board the Solar Dynamics Observatory (SDO) since its launch in 2010. In this paper, we look into another major instrument, namely the Michelson Doppler Imager (MDI) on board the Solar and Heliospheric Observatory (SOHO) from 1996 to 2010. The data archive of SOHO/MDI covers more active solar cycle 23 with many large flares. However, SOHO/MDI data only has line-of-sight (LOS) magnetograms. We propose a new deep learning method, named MagNet, to learn from combined LOS magnetograms, Bx and By taken by SDO/HMI along with H-alpha observations collected by the Big Bear Solar Observatory (BBSO), and to generate vector components Bx' and By', which would form vector magnetograms with observed LOS data. In this way, we can expand the availability of vector magnetograms to the period from 1996 to present. Experimental results demonstrate the good performance of the proposed method. To our knowledge, this is the first time that deep learning has been used to generate photospheric vector magnetograms of solar active regions for SOHO/MDI using SDO/HMI and H-alpha data.

astro-ph.SR↗

arXiv2022

VScript: Controllable Script Generation with Visual Presentation

In order to offer a customized script tool and inspire professional scriptwriters, we present VScript. It is a controllable pipeline that generates complete scripts, including dialogues and scene descriptions, as well as presents visually using video retrieval. With an interactive interface, our system allows users to select genres and input starting words that control the theme and development of the generated script. We adopt a hierarchical structure, which first generates the plot, then the script and its visual presentation. A novel approach is also introduced to plot-guided dialogue generation by treating it as an inverse dialogue summarization. The experiment results show that our approach outperforms the baselines on both automatic and human evaluations, especially in genre control.

cs.CL↗

arXiv2022

Inferring Line-of-Sight Velocities and Doppler Widths from Stokes Profiles of GST/NIRIS Using Stacked Deep Neural Networks

Obtaining high-quality magnetic and velocity fields through Stokes inversion is crucial in solar physics. In this paper, we present a new deep learning method, named Stacked Deep Neural Networks (SDNN), for inferring line-of-sight (LOS) velocities and Doppler widths from Stokes profiles collected by the Near InfraRed Imaging Spectropolarimeter (NIRIS) on the 1.6 m Goode Solar Telescope (GST) at the Big Bear Solar Observatory (BBSO). The training data of SDNN is prepared by a Milne-Eddington (ME) inversion code used by BBSO. We quantitatively assess SDNN, comparing its inversion results with those obtained by the ME inversion code and related machine learning (ML) algorithms such as multiple support vector regression, multilayer perceptrons and a pixel-level convolutional neural network. Major findings from our experimental study are summarized as follows. First, the SDNN-inferred LOS velocities are highly correlated to the ME-calculated ones with the Pearson product-moment correlation coefficient being close to 0.9 on average. Second, SDNN is faster, while producing smoother and cleaner LOS velocity and Doppler width maps, than the ME inversion code. Third, the maps produced by SDNN are closer to ME's maps than those from the related ML algorithms, demonstrating the better learning capability of SDNN than the ML algorithms. Finally, comparison between the inversion results of ME and SDNN based on GST/NIRIS and those from the Helioseismic and Magnetic Imager on board the Solar Dynamics Observatory in flare-prolific active region NOAA 12673 is presented. We also discuss extensions of SDNN for inferring vector magnetic fields with empirical evaluation.

astro-ph.SR↗

arXiv2022

A remark about time-analyticity of the linear Landau equation with soft potential

In this note, we study the Cauchy problem of the linear spatially homogeneous Landau equation with soft potentials. We prove that the solution to the Cauchy problem enjoys the analytic regularizing effect of the time variable with an L2 initial datum for positive time. So that the smoothing effect of Cauchy problem for the linear spatially homogeneous Landau equation with soft potentials is similar to the heat equation.

math.AP↗

arXiv2022

NeuralMarker: A Framework for Learning General Marker Correspondence

We tackle the problem of estimating correspondences from a general marker, such as a movie poster, to an image that captures such a marker. Conventionally, this problem is addressed by fitting a homography model based on sparse feature matching. However, they are only able to handle plane-like markers and the sparse features do not sufficiently utilize appearance information. In this paper, we propose a novel framework NeuralMarker, training a neural network estimating dense marker correspondences under various challenging conditions, such as marker deformation, harsh lighting, etc. Besides, we also propose a novel marker correspondence evaluation method circumstancing annotations on real marker-image pairs and create a new benchmark. We show that NeuralMarker significantly outperforms previous methods and enables new interesting applications, including Augmented Reality (AR) and video editing.

cs.CV↗

arXiv2022

Extended Cyclic Codes Sandwiched Between Reed-Muller Codes

The famous Barnes-Wall lattices can be obtained by applying Construction D to a chain of Reed-Muller codes. By applying Construction ${D}^{(cyc)}$ to a chain of extended cyclic codes sandwiched between Reed-Muller codes, Hu and Nebe (J. London Math. Soc. (2) 101 (2020) 1068-1089) constructed new series of universally strongly perfect lattices sandwiched between Barnes-Wall lattices. In this paper, we first extend their construction to generalized Reed-Muller codes, and then explicitly determine the minimum vectors of those new sandwiched Reed-Muller codes for some special cases.

cs.IT↗

arXiv2022

Iterative Adaptively Regularized LASSO-ADMM Algorithm for CFAR Estimation of Sparse Signals: IAR-LASSO-ADMM-CFAR Algorithm

The least-absolute shrinkage and selection operator (LASSO) is a regularization technique for estimating sparse signals of interest emerging in various applications and can be efficiently solved via the alternating direction method of multipliers (ADMM), which will be termed as LASSO-ADMM algorithm. The choice of the regularization parameter has significant impact on the performance of LASSO-ADMM algorithm. However, the optimization for the regularization parameter in the existing LASSO-ADMM algorithms has not been solved yet. In order to optimize this regularization parameter, we propose an efficient iterative adaptively regularized LASSO-ADMM (IAR-LASSO-ADMM) algorithm by iteratively updating the regularization parameter in the LASSO-ADMM algorithm. Moreover, a method is designed to iteratively update the regularization parameter by adding an outer iteration to the LASSO-ADMM algorithm. Specifically, at each outer iteration the zero support of the estimate obtained by the inner LASSO-ADMM algorithm is utilized to estimate the noise variance, and the noise variance is utilized to update the threshold according to a pre-defined const false alarm rate (CFAR). Then, the resulting threshold is utilized to update both the non-zero support of the estimate and the regularization parameter, and proceed to the next inner iteration. In addition, a suitable stopping criterion is designed to terminate the outer iteration process to obtain the final non-zero support of the estimate of the sparse measurement signals. The resulting algorithm is termed as IAR-LASSO-ADMM-CFAR algorithm. Finally, simulation results have been presented to show that the proposed IAR-LASSO-ADMM-CFAR algorithm outperforms the conventional LASSO-ADMM algorithm and other existing algorithms in terms of reconstruction accuracy, and its sparsity order estimate is more accurate than the existing algorithms.

cs.IT↗

arXiv2022

Nanoscale three-dimensional magnetic sensing with a probabilistic nanomagnet driven by spin-orbit torque

Detection of vector magnetic fields at nanoscale dimensions is critical in applications ranging from basic material science, to medical diagnostic. Meanwhile, an all-electric operation is of great significance for achieving a simple and compact sensing system. Here, we propose and experimentally demonstrate a simple approach to sensing a vector magnetic field at nanoscale dimensions, by monitoring a probabilistic nanomagnet's transition probability from a metastable state, excited by a driving current due to SOT, to a settled state. We achieve sensitivities for Hx, Hy, and Hz of 1.02%/Oe, 1.09%/Oe and 3.43%/Oe, respectively, with a 200 x 200 nm^2 nanomagnet. The minimum detectable field is dependent on the driving pulse events N, and is expected to be as low as 1 uT if N = 3 x 10^6.

cond-mat.mtrl-sci↗

arXiv2022

Structure-preserving finite volume arbitrary Lagrangian-Eulerian WENO schemes for the shallow water equations

This paper develops the structure-preserving finite volume weighted essentially non-oscillatory (WENO) hybrid schemes for the shallow water equations under the arbitrary Lagrangian-Eulerian (ALE) framework, dubbed as ALE-WENO schemes. The WENO hybrid reconstruction is adopted on moving meshes, which distinguishes the smooth, non-smooth, and transition stencils by a simple smoothness detector. To maintain the positivity preserving and the well-balanced properties of the ALE-WENO schemes, we adapt the positivity preserving limiter and the well-balanced approaches on static meshes to moving meshes. The rigorous theoretical analysis and numerical examples demonstrate the high order accuracy and positivity-preserving property of the schemes under the ALE framework. For the well-balanced schemes, it is successful in the unique exact equilibrium preservation and capturing small perturbations of the hydrostatic state well without numerical oscillations near the discontinuity. Moreover, our ALE-WENO hybrid schemes have an advantage over the simulations on static meshes due to the higher resolution interface tracking of the fluid motion.

math.NA↗

arXiv2022

LAMOST meets Gaia: The Galactic Open Clusters

Open Clusters are born and evolve along the Milky Way plane, on them is imprinted the history of the Galactic disc, including the chemical and dynamical evolution. Chemical and dynamical properties of open clusters can be derived from photometric, spectroscopic, and astrometric data of their member stars. Based on the photometric and astrometric data from the Gaia mission, the membership of stars in more than 2000 Galactic clusters has been identified in the literature. The chemical and kinematical properties, however, are still poorly known for many of these clusters. In synergy with the large spectroscopic survey LAMOST (data release 8) and Gaia (data release 2), we report a new comprehensive catalogue of 386 open clusters. This catalogue has homogeneous parameter determinations of radial velocity, metallicity, and dynamical properties, such as orbit, eccentricity, angular momenta, total energy, and 3D Galactic velocity. These parameters allow the first radial velocity determination and the first spectroscopic [Fe/H] determination for 44 and 137 clusters, respectively. The metallicity distribution of majority clusters shows falling trends in the parameter space of the Galactocentric radius, the total energy, and the Z component of angular momentum -- except for two old groups that show flat tails in their own parameter planes. Cluster populations of ages younger and older than 500 Myrs distribute diversely on the disc. The latter has a spatial consistency with the Galactic disc flare. The 3-D spatial comparison between very young clusters (< 100 Myr) and nearby molecular clouds revealed a wide range of metallicity distribution along the Radcliffe gas cloud wave, indicating a possible inhomogeneous mixing or fast star formation along the wave. This catalogue would serve the community as a useful tool to trace the chemical and dynamical evolution of the Milky Way.

astro-ph.GA↗

arXiv2022

Improving Visual Grounding with Visual-Linguistic Verification and Iterative Reasoning

Visual grounding is a task to locate the target indicated by a natural language expression. Existing methods extend the generic object detection framework to this problem. They base the visual grounding on the features from pre-generated proposals or anchors, and fuse these features with the text embeddings to locate the target mentioned by the text. However, modeling the visual features from these predefined locations may fail to fully exploit the visual context and attribute information in the text query, which limits their performance. In this paper, we propose a transformer-based framework for accurate visual grounding by establishing text-conditioned discriminative features and performing multi-stage cross-modal reasoning. Specifically, we develop a visual-linguistic verification module to focus the visual features on regions relevant to the textual descriptions while suppressing the unrelated areas. A language-guided feature encoder is also devised to aggregate the visual contexts of the target object to improve the object's distinctiveness. To retrieve the target from the encoded visual features, we further propose a multi-stage cross-modal decoder to iteratively speculate on the correlations between the image and text for accurate target localization. Extensive experiments on five widely used datasets validate the efficacy of our proposed components and demonstrate state-of-the-art performance. Our code is public at https://github.com/yangli18/VLTVG.

cs.CV↗

arXiv2022

Analytic smoothing effect of the time variable for the spatially homogeneous Landau equation

In this work, we study the Cauchy problem of the spatially homogeneous Landau equation with hard potentials in a close-to-quilibrium framework. We prove that the solution to the Cauchy problem enjoys the analytic regularizing effect of the time variable with an L2 initial datum for positive time. So that the smoothing effect of Cauchy problem for the spatially homogeneous Landau equation with hard potentials is exactly same as heat equation.

math.AP↗

arXiv2022

Transformer based multiple instance learning for weakly supervised histopathology image segmentation

Hispathological image segmentation algorithms play a critical role in computer aided diagnosis technology. The development of weakly supervised segmentation algorithm alleviates the problem of medical image annotation that it is time-consuming and labor-intensive. As a subset of weakly supervised learning, Multiple Instance Learning (MIL) has been proven to be effective in segmentation. However, there is a lack of related information between instances in MIL, which limits the further improvement of segmentation performance. In this paper, we propose a novel weakly supervised method for pixel-level segmentation in histopathology images, which introduces Transformer into the MIL framework to capture global or long-range dependencies. The multi-head self-attention in the Transformer establishes the relationship between instances, which solves the shortcoming that instances are independent of each other in MIL. In addition, deep supervision is introduced to overcome the limitation of annotations in weakly supervised methods and make the better utilization of hierarchical information. The state-of-the-art results on the colon cancer dataset demonstrate the superiority of the proposed method compared with other weakly supervised methods. It is worth believing that there is a potential of our approach for various applications in medical images.

cs.CV↗

arXiv2022

3D Segmentation Guided Style-based Generative Adversarial Networks for PET Synthesis

Potential radioactive hazards in full-dose positron emission tomography (PET) imaging remain a concern, whereas the quality of low-dose images is never desirable for clinical use. So it is of great interest to translate low-dose PET images into full-dose. Previous studies based on deep learning methods usually directly extract hierarchical features for reconstruction. We notice that the importance of each feature is different and they should be weighted dissimilarly so that tiny information can be captured by the neural network. Furthermore, the synthesis on some regions of interest is important in some applications. Here we propose a novel segmentation guided style-based generative adversarial network (SGSGAN) for PET synthesis. (1) We put forward a style-based generator employing style modulation, which specifically controls the hierarchical features in the translation process, to generate images with more realistic textures. (2) We adopt a task-driven strategy that couples a segmentation task with a generative adversarial network (GAN) framework to improve the translation performance. Extensive experiments show the superiority of our overall framework in PET synthesis, especially on those regions of interest.

eess.IV↗

arXiv2022

Retrieval-Free Knowledge-Grounded Dialogue Response Generation with Adapters

To diversify and enrich generated dialogue responses, knowledge-grounded dialogue has been investigated in recent years. The existing methods tackle the knowledge grounding challenge by retrieving the relevant sentences over a large corpus and augmenting the dialogues with explicit extra information. Despite their success, however, the existing works have drawbacks in inference efficiency. This paper proposes KnowExpert, a framework to bypass the explicit retrieval process and inject knowledge into the pre-trained language models with lightweight adapters and adapt to the knowledge-grounded dialogue task. To the best of our knowledge, this is the first attempt to tackle this challenge without retrieval in this task under an open-domain chit-chat scenario. The experimental results show that Knowexpert performs comparably with some retrieval-based baselines while being time-efficient in inference, demonstrating the effectiveness of our proposed method.

cs.CL↗