arXiv subjects

Xu Yang

Publications and source records attributed to Xu Yang.

At least 343 records · Page 19Linked to original sources

arXiv2021

Auto-Parsing Network for Image Captioning and Visual Question Answering

We propose an Auto-Parsing Network (APN) to discover and exploit the input data's hidden tree structures for improving the effectiveness of the Transformer-based vision-language systems. Specifically, we impose a Probabilistic Graphical Model (PGM) parameterized by the attention operations on each self-attention layer to incorporate sparse assumption. We use this PGM to softly segment an input sequence into a few clusters where each cluster can be treated as the parent of the inside entities. By stacking these PGM constrained self-attention layers, the clusters in a lower layer compose into a new sequence, and the PGM in a higher layer will further segment this sequence. Iteratively, a sparse tree can be implicitly parsed, and this tree's hierarchical knowledge is incorporated into the transformed embeddings, which can be used for solving the target vision-language tasks. Specifically, we showcase that our APN can strengthen Transformer based networks in two major vision-language tasks: Captioning and Visual Question Answering. Also, a PGM probability-based parsing algorithm is developed by which we can discover what the hidden structure of input is during the inference.

cs.CV↗

arXiv2021

Stripe order in the doped Hubbard model on the honeycomb lattice

We study the ground state properties of the doped Hubbard model with strong interactions on honeycomb lattice by the Density Matrix Renormalization Group (DMRG) method. At half-filling, due to the absence of minus sign problem, it is now well established by large-scale Quantum Monte Carlo calculations that a Dirac semi-metal to anti-ferromagnetic Mott insulator transition occurs with the increase of the interaction strength $U$ for the Hubbard model on honeycomb lattice. However, an understanding of the fate of the anti-ferromagnetic Mott insulator when holes are doped into the system is still lacking. In this work, by calculating the local spin and charge density for width-4 cylinders with DMRG, we discover a half-filled stripe order in the doped Hubbard model on honeycomb lattice. We also perform complementary large-scale mean-field calculations with renormalized interaction strength. We observe half-filled stripe order and find stripe states with filling close to one half are nearly degenerate in energy.

cond-mat.str-el↗

arXiv2021

Well-posedness of martingale problem for SBM with interacting branching

In this paper a martingale problem for super-Brownian motion with interactive branching is derived. The uniqueness of the solution to the martingale problem is obtained by using the pathwise uniqueness of the solution to a corresponding system of SPDEs with proper boundary conditions. The existence of the solution to the martingale problem and the Holder continuity of the density process are also studied.

math.PR↗

arXiv2021

Doubly Contrastive Deep Clustering

Deep clustering successfully provides more effective features than conventional ones and thus becomes an important technique in current unsupervised learning. However, most deep clustering methods ignore the vital positive and negative pairs introduced by data augmentation and further the significance of contrastive learning, which leads to suboptimal performance. In this paper, we present a novel Doubly Contrastive Deep Clustering (DCDC) framework, which constructs contrastive loss over both sample and class views to obtain more discriminative features and competitive results. Specifically, for the sample view, we set the class distribution of the original sample and its augmented version as positive sample pairs and set one of the other augmented samples as negative sample pairs. After that, we can adopt the sample-wise contrastive loss to pull positive sample pairs together and push negative sample pairs apart. Similarly, for the class view, we build the positive and negative pairs from the sample distribution of the class. In this way, two contrastive losses successfully constrain the clustering results of mini-batch samples in both sample and class level. Extensive experimental results on six benchmark datasets demonstrate the superiority of our proposed model against state-of-the-art methods. Particularly in the challenging dataset Tiny-ImageNet, our method leads 5.6\% against the latest comparison method. Our code will be available at \url{https://github.com/ZhiyuanDang/DCDC}.

cs.CV↗

arXiv2021

Causal Attention for Vision-Language Tasks

We present a novel attention mechanism: Causal Attention (CATT), to remove the ever-elusive confounding effect in existing attention-based vision-language models. This effect causes harmful bias that misleads the attention module to focus on the spurious correlations in training data, damaging the model generalization. As the confounder is unobserved in general, we use the front-door adjustment to realize the causal intervention, which does not require any knowledge on the confounder. Specifically, CATT is implemented as a combination of 1) In-Sample Attention (IS-ATT) and 2) Cross-Sample Attention (CS-ATT), where the latter forcibly brings other samples into every IS-ATT, mimicking the causal intervention. CATT abides by the Q-K-V convention and hence can replace any attention module such as top-down attention and self-attention in Transformers. CATT improves various popular attention-based vision-language models by considerable margins. In particular, we show that CATT has great potential in large-scale pre-training, e.g., it can promote the lighter LXMERT~\cite{tan2019lxmert}, which uses fewer data and less computational power, comparable to the heavier UNITER~\cite{chen2020uniter}. Code is published in \url{https://github.com/yangxuntu/catt}.

cs.CV↗

arXiv2021

Classical limit for the varying-mass Schrödinger equation with random inhomogeneities

The varying-mass Schrödinger equation (VMSE) has been successfully applied to model electronic properties of semiconductor hetero-structures, for example, quantum dots and quantum wells. In this paper, we consider VMSE with small random heterogeneities, and derive a radiative transfer equation as its asymptotic limit. The main tool is to systematically apply the Wigner transform in the classical regime when the rescaled Planck constant $ε\ll 1$, and expand the Wigner equation to proper orders of $ε$. As a proof of concept, we numerically compute both VMSE and its limiting radiative transfer equation, and show that their solutions agree well in the classical regime.

math.NA↗

arXiv2021

Cloud Cover and Aurora Contamination at Dome A in 2017 from KLCAM

Dome A in Antarctica has many characteristics that make it an excellent site for astronomical observations, from the optical to the terahertz. Quantitative site testing is still needed to confirm the site's properties. In this paper, we present a statistical analysis of cloud cover and aurora contamination from the Kunlun Cloud and Aurora Monitor (KLCAM). KLCAM is an automatic, unattended all-sky camera aiming for long-term monitoring of the usable observing time and optical sky background at Dome~A. It was installed at Dome~A in January 2017, worked through the austral winter, and collected over 47,000 images over 490 days. A semi-quantitative visual data analysis of cloud cover and auroral contamination was carried out by five individuals. The analysis shows that the night sky was free of cloud for 83 per cent of the time, which ranks Dome~A highly in a comparison with other observatory sites. Although aurorae were detected somewhere on an image for nearly 45 per cent of the time, the strongest auroral emission lines can be filtered out with customized filters.

astro-ph.IM↗

arXiv2020

Incremental Embedding Learning via Zero-Shot Translation

Modern deep learning methods have achieved great success in machine learning and computer vision fields by learning a set of pre-defined datasets. Howerver, these methods perform unsatisfactorily when applied into real-world situations. The reason of this phenomenon is that learning new tasks leads the trained model quickly forget the knowledge of old tasks, which is referred to as catastrophic forgetting. Current state-of-the-art incremental learning methods tackle catastrophic forgetting problem in traditional classification networks and ignore the problem existing in embedding networks, which are the basic networks for image retrieval, face recognition, zero-shot learning, etc. Different from traditional incremental classification networks, the semantic gap between the embedding spaces of two adjacent tasks is the main challenge for embedding networks under incremental learning setting. Thus, we propose a novel class-incremental method for embedding network, named as zero-shot translation class-incremental method (ZSTCI), which leverages zero-shot translation to estimate and compensate the semantic gap without any exemplars. Then, we try to learn a unified representation for two adjacent tasks in sequential learning process, which captures the relationships of previous classes and current classes precisely. In addition, ZSTCI can easily be combined with existing regularization-based incremental learning methods to further improve performance of embedding networks. We conduct extensive experiments on CUB-200-2011 and CIFAR100, and the experiment results prove the effectiveness of our method. The code of our method has been released.

cs.CV↗

arXiv2020

Boundary behaviors for a class of continuous-state nonlinear branching processes in critical cases

Using Foster-Lyapunov techniques we establish new conditions on non-extinction, non-explosion, coming down from infinity and staying infinite, respectively, for the general continuous-state nonlinear branching processes introduced in Li et al. (2019). These results can be applied to identify boundary behaviors for the critical cases of the above nonlinear branching processes with power rate functions driven by Brownian motion and (or) stable Poisson random measure, which was left open in Li et al. (2019). In particular, we show that even in the critical cases, a phase transition happens between coming down from infinity and staying infinite.

math.PR↗

arXiv2020

Discovery of segmented Fermi surface induced by Cooper pair momentum

Since the early days of Bardeen-Cooper-Schrieffer theory, it has been predicted that a sufficiently large supercurrent can close the energy gap in a superconductor and creates gapless Bogoliubov quasiparticles through the Doppler shift of quasiparticle energy due to the Cooper pair momentum. In this gapless superconducting state, zero-energy quasiparticles reside on a segment of the normal state Fermi surface, while its remaining part is still gapped. The finite density of states of field-induced quasiparticles, known as the Volovik effect, has been observed in tunneling and specific heat measurements on d- and s-wave superconductors. However, the segmented Fermi surface of a finite-momentum state carrying a supercurrent has never been detected directly. Here we use quasiparticle interference (QPI) technique to image field-controlled Fermi surface of Bi$_2$Te$_3$ thin films proximitized by the superconductor NbSe$_2$. By applying a small in-plane magnetic field, a screening supercurrent is induced which leads to finite-momentum pairing on topological surface states of Bi$_2$Te$_3$. Our measurements and analysis reveal the strong impact of finite Cooper pair momentum on the quasiparticle spectrum, and thus pave the way for STM study of pair density wave and FFLO states in unconventional superconductors.

cond-mat.supr-con↗

arXiv2020

Finding It at Another Side: A Viewpoint-Adapted Matching Encoder for Change Captioning

Change Captioning is a task that aims to describe the difference between images with natural language. Most existing methods treat this problem as a difference judgment without the existence of distractors, such as viewpoint changes. However, in practice, viewpoint changes happen often and can overwhelm the semantic difference to be described. In this paper, we propose a novel visual encoder to explicitly distinguish viewpoint changes from semantic changes in the change captioning task. Moreover, we further simulate the attention preference of humans and propose a novel reinforcement learning process to fine-tune the attention directly with language evaluation rewards. Extensive experimental results show that our method outperforms the state-of-the-art approaches by a large margin in both Spot-the-Diff and CLEVR-Change datasets.

cs.CV↗

arXiv2020

CYRA: the cryogenic infrared spectrograph for the Goode Solar Telescope in Big Bear

CYRA (CrYogenic solar spectrogRAph) is a facility instrument of the 1.6-meter Goode Solar Telescope (GST) at the Big Bear Solar Observatory (BBSO). CYRA focuses on the study of the near-infrared solar spectrum between 1 and 5 microns, a under explored region which is not only a fertile ground for photospheric magnetic diagnostics, but also allows a unique window into the chromosphere lying atop the photosphere. CYRA is the first ever fully cryogenic spectrograph in any solar observatory with its two predecessors, on the McMath-Pierce and Mees Telescopes, being based on warm optics except for the detectors and order sorting filters. CYRA is used to probe magnetic fields in various solar features and the quiet photosphere. CYRA measurements will allow new and better 3D extrapolations of the solar magnetic field and will provide more accurate boundary conditions for solar activity models. Superior spectral resolution of 150,000 and better allows enhanced observations of the chromosphere in the carbon monoxide (CO) spectral bands and will yield a better understanding of energy transport in the solar atmosphere. CYRA is divided into two optical sub-systems: The Fore-Optics Module and the Spectrograph. The Spectrograph is the heart of the instrument and contains the IR detector, grating, slits, filters, and imaging optics all in a cryogenically cooled Dewar (cryostat). The detector a 2048 by 2048 pixel HAWAII 2 array produced by Teledyne Scientific & Imaging, LLC. The interior of the cryostat and the readout electronics are maintained at 90 Kelvin by helium refrigerant based cryo-coolers, while the IR array is cooled to 30 Kelvin. The Fore-Optics Module de-rotates and stabilizes the solar image, provides scanning capabilities, and transfers the GST image to the Spectrograph. CYRA has been installed and is undergoing its commissioning phase.

astro-ph.SR↗

arXiv2020

Night-time measurements of astronomical seeing at Dome A in Antarctica

Seeing, the angular size of stellar images blurred by atmospheric turbulence, is a critical parameter used to assess the quality of astronomical sites. Median values at the best mid-latitude sites are generally in the range of 0.6--0.8\,arcsec. Sites on the Antarctic plateau are characterized by comparatively-weak turbulence in the free-atmosphere above a strong but thin boundary layer. The median seeing at Dome C is estimated to be 0.23--0.36 arcsec above a boundary layer that has a typical height of 30\,m. At Dome A and F, the only previous seeing measurements were made during daytime. Here we report the first direct measurements of night-time seeing at Dome A, using a Differential Image Motion Monitor. Located at a height of just 8\,m, it recorded seeing as low as 0.13\,arcsec, and provided seeing statistics that are comparable to those for a 20\,m height at Dome C. It indicates that the boundary layer was below 8\,m 31\% of the time. At such times the median seeing was 0.31\,arcsec, consistent with free-atmosphere seeing. The seeing and boundary layer thickness are found to be strongly correlated with the near-surface temperature gradient. The correlation confirms a median thickness of approximately 14\,m for the boundary layer at Dome A, as found from a sonic radar. The thinner boundary layer makes it less challenging to locate a telescope above it, thereby giving greater access to the free-atmosphere.

astro-ph.IM↗

arXiv2020

Unfitted Nitsche's method for computing band structures in phononic crystals with impurities

In this paper, we propose an unfitted Nitsche's method to compute the band structures of phononic crystal with impurities of general geometry. The proposed method does not require the background mesh to fit the interfaces of impurities, and thus avoids the expensive cost of generating body-fitted meshes and simplifies the inclusion of interface conditions in the formulation. The quasi-periodic boundary conditions are handled by the Floquet-Bloch transform, which converts the computation of band structures into an eigenvalue problem with periodic boundary conditions. More importantly, we show the well-posedness of the proposed method using a delicate argument based on the trace inequality, and further prove the convergence by the Babuška-Osborn theory. We achieve the optimal convergence rate at the presence of the impurities of general geometry. We confirm the theoretical results by two numerical examples, and show the capability of the proposed methods for computing the band structures without fitting the interfaces of impurities.

math.NA↗

arXiv2020

Large spin to charge conversion in topological superconductor \b{eta}-PdBi2 at room temperature

\b{eta}-PdBi2 has attracted much attention for its prospective ability to possess simultaneously topological surface and superconducting states due to its unprecedented spin-orbit interaction (SOC). Whereas most works have focused solely on investigating its topological surface states, the coupling between spin and charge degrees of freedom in this class of quantum material remains unexplored. Here we first report a study of spin-to-charge conversion in a \b{eta}-PdBi2 ultrathin film grown by molecular beam epitaxy, utilizing a spin pumping technique to perform inverse spin Hall effect measurements. We find that the room temperature spin Hall angle of Fe/\b{eta}-PdBi2, θ_SH=0.037. This value is one order of magnitude larger than that of reported conventional superconductors, and is comparable to that of the best SOC metals and topological insulators. Our results provide an avenue for developing superconductor-based spintronic applications.

cond-mat.mes-hall↗

arXiv2020

Automated Pavement Crack Segmentation Using U-Net-based Convolutional Neural Network

Automated pavement crack image segmentation is challenging because of inherent irregular patterns, lighting conditions, and noise in images. Conventional approaches require a substantial amount of feature engineering to differentiate crack regions from non-affected regions. In this paper, we propose a deep learning technique based on a convolutional neural network to perform segmentation tasks on pavement crack images. Our approach requires minimal feature engineering compared to other machine learning techniques. We propose a U-Net-based network architecture in which we replace the encoder with a pretrained ResNet-34 neural network. We use a "one-cycle" training schedule based on cyclical learning rates to speed up the convergence. Our method achieves an F1 score of 96% on the CFD dataset and 73% on the Crack500 dataset, outperforming other algorithms tested on these datasets. We perform ablation studies on various techniques that helped us get marginal performance boosts, i.e., the addition of spatial and channel squeeze and excitation (SCSE) modules, training with gradually increasing image sizes, and training various neural network layers with different learning rates.

cs.CV↗

arXiv2020

Unfitted Nitsche's method for computing wave modes in topological materials

In this paper, we propose an unfitted Nitsche's method for computing wave modes in topological materials. The proposed method is based on Nitsche's technique to study the performance-enhanced topological materials which have strongly heterogeneous structures (e.g., the refractive index is piecewise constant with high contrasts). For periodic bulk materials, we use Floquet-Bloch theory and solve an eigenvalue problem on a torus with unfitted meshes. For the materials with a line defect, a sufficiently large domain with zero boundary conditions is used to compute the localized eigenfunctions corresponding to the edge modes. The interfaces are handled by Nitsche's method on an unfitted uniform mesh. We prove the proposed methods converge optimally, and present numerical examples to validate the theoretical results and demonstrate the capability of simulating topological materials.

math.NA↗

arXiv2020

Automation of the AST3 optical sky survey from Dome~A, Antarctica

The 0.5\,m Antarctic Survey Telescopes (AST3) were designed for time-domain optical/infrared astronomy. They are located in Dome~A, Antarctica, where they can take advantage of the continuous dark time during winter. Since the site is unattended in winter, everything for the operation, from observing to data reduction, had to be fully automated. Here, we present a brief overview of the AST3 project and some of its unique characteristics due to its location in Antarctica. We summarise the various components of the survey, including the customized hardware and software, that make complete automation possible.

astro-ph.IM↗