arXiv subjects

Chenyu Wang

Publications and source records attributed to Chenyu Wang.

124 records · Page 7Linked to original sources

arXiv2022

Decompose to Adapt: Cross-domain Object Detection via Feature Disentanglement

Recent advances in unsupervised domain adaptation (UDA) techniques have witnessed great success in cross-domain computer vision tasks, enhancing the generalization ability of data-driven deep learning architectures by bridging the domain distribution gaps. For the UDA-based cross-domain object detection methods, the majority of them alleviate the domain bias by inducing the domain-invariant feature generation via adversarial learning strategy. However, their domain discriminators have limited classification ability due to the unstable adversarial training process. Therefore, the extracted features induced by them cannot be perfectly domain-invariant and still contain domain-private factors, bringing obstacles to further alleviate the cross-domain discrepancy. To tackle this issue, we design a Domain Disentanglement Faster-RCNN (DDF) to eliminate the source-specific information in the features for detection task learning. Our DDF method facilitates the feature disentanglement at the global and local stages, with a Global Triplet Disentanglement (GTD) module and an Instance Similarity Disentanglement (ISD) module, respectively. By outperforming state-of-the-art methods on four benchmark UDA object detection tasks, our DDF method is demonstrated to be effective with wide applicability.

cs.CV↗

arXiv2021

Early Stage Growth of Amorphous Thin Film: Average Kinetics, Nanoscale Dynamics and Pressure Dependence

We used the Coherent Grazing Incidence Small Angle X-Ray Scattering (Co-GISAXS) technique to study the average kinetics and nanoscale dynamics during early-stage a-WSi$_2$ sputter deposition. The kinetic and dynamic properties are examined as a function of pressure, which is known to be a critical factor in determining final surface roughness. Surface growth kinetics and dynamics are characterized by time parameters extracted from the height-height structure factor and correlation functions. The roughness at a given length scale reaches a maximum before relaxing down to a steady state. The lateral length scale dependence and pressure dependence are then compared among measured kinetics and dynamics time parameters. Surfaces grown at lower pressures are smoother, leading to longer correlation times. The time to reach a dynamic steady state and a kinetic steady state show contrasting pressure dependence. A dynamic steady state is reached earlier than the kinetic steady state at high pressure. A more random deposition direction and lower kinetic energy at higher pressures can explain these phenomena, along with the hypothesis that larger nanoclusters form in vapor before arriving at the surface. A continuum model is applied to simulate the overall behavior with mixed success.

cond-mat.mtrl-sci↗

arXiv2021

HAGEN: Homophily-Aware Graph Convolutional Recurrent Network for Crime Forecasting

The crime forecasting is an important problem as it greatly contributes to urban safety. Typically, the goal of the problem is to predict different types of crimes for each geographical region (like a neighborhood or censor tract) in the near future. Since nearby regions usually have similar socioeconomic characteristics which indicate similar crime patterns, recent state-of-the-art solutions constructed a distance-based region graph and utilized Graph Neural Network (GNN) techniques for crime forecasting, because the GNN techniques could effectively exploit the latent relationships between neighboring region nodes in the graph. However, this distance-based pre-defined graph cannot fully capture crime correlation between regions that are far from each other but share similar crime patterns. Hence, to make an accurate crime prediction, the main challenge is to learn a better graph that reveals the dependencies between regions in crime occurrences and meanwhile captures the temporal patterns from historical crime records. To address these challenges, we propose an end-to-end graph convolutional recurrent network called HAGEN with several novel designs for crime prediction. Specifically, our framework could jointly capture the crime correlation between regions and the temporal crime dynamics by combining an adaptive region graph learning module with the Diffusion Convolution Gated Recurrent Unit (DCGRU). Based on the homophily assumption of GNN, we propose a homophily-aware constraint to regularize the optimization of the region graph so that neighboring region nodes on the learned graph share similar crime patterns, thus fitting the mechanism of diffusion convolution. It also incorporates crime embedding to model the interdependencies between regions and crime categories. Empirical experiments and comprehensive analysis on two real-world datasets showcase the effectiveness of HAGEN.

cs.LG↗

arXiv2021

SPI: Automated Identification of Security Patches via Commits

Security patches in open-source software, providing security fixes to identified vulnerabilities, are crucial in protecting against cyberattacks. Despite the National Vulnerability Database (NVD) publishes identified vulnerabilities, a vast majority of vulnerabilities and their corresponding security patches remain beyond public exposure, e.g., in the open-source libraries that are heavily relied on by developers. An extensive security patches dataset could help end-users such as security companies, e.g., building a security knowledge base, or researchers, e.g., aiding in vulnerability research. To curate security patches including undisclosed patches at a large scale and low cost, we propose a deep neural-network-based approach built upon commits of open-source repositories. We build security patch datasets that include 38,291 security-related commits and 1,045 CVE patches from four C libraries. We manually verify each commit, among the 38,291 security-related commits, to determine if they are security-related. We devise a deep learning-based security patch identification system that consists of two neural networks: one commit-message neural network that utilizes pretrained word representations learned from our commits dataset; and one code-revision neural network that takes code before and after revision and learns the distinction on the statement level. Our evaluation results show that our system outperforms SVM and K-fold stacking algorithm, achieving as high as 87.93% F1-score and precision of 86.24%. We deployed our pipeline and learned model in an industrial production environment to evaluate the generalization ability of our approach. The industrial dataset consists of 298,917 commits from 410 new libraries that range from a wide functionality. Our experiment results and observation proved that our approach identifies security patches effectively among open-sourced projects.

cs.CR↗

arXiv2021

Temperature-sensitive spatial distribution of defects in PdSe2 flakes

Defect engineering plays an important role in tailoring the electronic transport properties of van der Waals materials. However, it is usually achieved through tuning the type and concentration of defects, rather than dynamically reconfiguring their spatial distribution. Here, we report temperature-sensitive spatial redistribution of defects in PdSe2 thin flakes through scanning tunneling microscopy (STM). We observe that the spatial distribution of Se vacancies in PdSe2 flakes exhibits a strong anisotropic characteristic at 80 K, and that this orientation-dependent feature is weakened when temperature is raised. Moreover, we carry out transport measurements on PdSe2 thin flakes and show that the anisotropic features of carrier mobility and phase coherent length are also sensitive to temperature. Combining with theoretical analysis, we conclude that temperature-driven defect spatial redistribution could interpret the temperature-sensitive electrical transport behaviors in PdSe2 thin flakes. Our work highlights that engineering spatial distribution of defects in the van der Waals materials, which has been overlooked before, may open up a new avenue to tailor the physical properties of materials and explore new device functionalities.

cond-mat.mes-hall↗

arXiv2021

Open Domain Generalization with Domain-Augmented Meta-Learning

Leveraging datasets available to learn a model with high generalization ability to unseen domains is important for computer vision, especially when the unseen domain's annotated data are unavailable. We study a novel and practical problem of Open Domain Generalization (OpenDG), which learns from different source domains to achieve high performance on an unknown target domain, where the distributions and label sets of each individual source domain and the target domain can be different. The problem can be generally applied to diverse source domains and widely applicable to real-world applications. We propose a Domain-Augmented Meta-Learning framework to learn open-domain generalizable representations. We augment domains on both feature-level by a new Dirichlet mixup and label-level by distilled soft-labeling, which complements each domain with missing classes and other domain knowledge. We conduct meta-learning over domains by designing new meta-learning tasks and losses to preserve domain unique knowledge and generalize knowledge across domains simultaneously. Experiment results on various multi-domain datasets demonstrate that the proposed Domain-Augmented Meta-Learning (DAML) outperforms prior methods for unseen domain recognition.

cs.CV↗

arXiv2019

Course Concept Expansion in MOOCs with External Knowledge and Interactive Game

As Massive Open Online Courses (MOOCs) become increasingly popular, it is promising to automatically provide extracurricular knowledge for MOOC users. Suffering from semantic drifts and lack of knowledge guidance, existing methods can not effectively expand course concepts in complex MOOC environments. In this paper, we first build a novel boundary during searching for new concepts via external knowledge base and then utilize heterogeneous features to verify the high-quality results. In addition, to involve human efforts in our model, we design an interactive optimization mechanism based on a game. Our experiments on the four datasets from Coursera and XuetangX show that the proposed method achieves significant improvements(+0.19 by MAP) over existing methods. The source code and datasets have been published.

cs.CL↗

arXiv2019

Multiple Sclerosis Lesion Inpainting Using Non-Local Partial Convolutions

Multiple sclerosis (MS) is an inflammatory demyelinating disease of the central nervous system (CNS) that results in focal injury to the grey and white matter. The presence of white matter lesions biases morphometric analyses such as registration, individual longitudinal measurements and tissue segmentation for brain volume measurements. Lesion-inpainting with intensities derived from surrounding healthy tissue represents one approach to alleviate such problems. However, existing methods inpaint lesions based on texture information derived from local surrounding tissue, often leading to inconsistent inpainting and the generation of artifacts such as intensity discrepancy and blurriness. Based on these observations, we propose non-local partial convolutions (NLPC) that integrates a Unet-like network with the non-local module. The non-local module is exploited to capture long range dependencies between the lesion area and remaining normal-appearing brain regions. Then, the lesion area is filled by referring to normal-appearing regions with more similar features. This method generates inpainted regions that appear more realistic and natural. Our quantitative experimental results also demonstrate superiority of this technique of existing state-of-the-art inpainting methods.

cs.CV↗

arXiv2018

Negative Photoconductance in van der Waals Heterostructures-Based Floating Gate Phototransistor

Van der Waals (vdW) heterostructures made of two-dimensional materials have been demonstrated to be versatile architectures for optoelectronic applications due to strong light-matter interactions. However, most of light-controlled phenomena and applications in the vdW heterostructures rely on positive photoconductance (PPC). Negative photoconductance (NPC) has not yet been reported in vdW heterostructures. Here we report the observation of the NPC in ReS2/h-BN/MoS2 vdW heterostructures-based floating gate phototransistor. The fabricated devices exhibit excellent performance of nonvolatile memory without light illumination. More interestingly, we observe a gate-tunable transition between the PPC and the NPC under the light illumination. The observed NPC phenomenon can be attributed to the charge transfer between floating gate and conduction channel. Furthermore, we show that the control of NPC through light intensity is promising in realization of light-tunable multi-bit memory devices. Our results may enable potential applications in multifunctional memories and optoelectronic devices.

cond-mat.mes-hall↗

arXiv2018

Low-Temperature Eutectic Synthesis of PtTe2 with Weak Antilocalization and Controlled Layer Thinning

Metallic transition metal dichalcogenides (TMDs) have exhibited various exotic physical properties and hold the promise of novel optoelectronic and topological devices applications. However, the synthesis of metallic TMDs is based on gas-phase methods and requires high temperature condition. As an alternative to the gas-phase synthetic approach, lower temperature eutectic liquid-phase synthesis presents a very promising approach with the potential for larger-scale and controllable growth of high-quality thin metallic TMDs single crystals. Herein, we report the first realization of low-temperature eutectic liquid-phase synthesis of type-II Dirac semimetal PtTe2 single crystals with thickness ranging from 2 to 200 nm. The electrical measurement of synthesized PtTe2 reveals a record-high conductivity of as high as 3.3*106 S/m at room temperature. Besides, we experimentally identify the weak antilocalization behavior in the type-II Dirac semimetal PtTe2 for the first time. Furthermore, we develop a simple and general strategy to obtain atomically-thin PtTe2 crystal by thinning as-synthesized bulk samples, which can still retain highly crystalline and exhibits excellent electric conductivity. Our results of controllable and scalable low-temperature eutectic liquid-phase synthesis and layer-by-layer thinning of high-quality thin PtTe2 single crystals offer a simple and general approach for obtaining different thickness metallic TMDs with high-melting point transition metal.

cond-mat.mtrl-sci↗

arXiv2018

Gate-Induced Interfacial Superconductivity in 1T-SnSe2

Layered metal chalcogenide materials provide a versatile platform to investigate emergent phenomena and two-dimensional (2D) superconductivity at/near the atomically thin limit. In particular, gate-induced interfacial superconductivity realized by the use of an electric-double-layer transistor (EDLT) has greatly extended the capability to electrically induce superconductivity in oxides, nitrides and transition metal chalcogenides and enable one to explore new physics, such as the Ising pairing mechanism. Exploiting gate-induced superconductivity in various materials can provide us with additional platforms to understand emergent interfacial superconductivity. Here, we report the discovery of gate-induced 2D superconductivity in layered 1T-SnSe2, a typical member of the main-group metal dichalcogenide (MDC) family, using an EDLT gating geometry. A superconducting transition temperature Tc around 3.9 K was demonstrated at the EDL interface. The 2D nature of the superconductivity therein was further confirmed based on 1) a 2D Tinkham description of the angle-dependent upper critical field, 2) the existence of a quantum creep state as well as a large ratio of the coherence length to the thickness of superconductivity. Interestingly, the in-plane approaching zero temperature was found to be 2-3 times higher than the Pauli limit, which might be related to an electric field-modulated spin-orbit interaction. Such results provide a new perspective to expand the material matrix available for gate-induced 2D superconductivity and the fundamental understanding of interfacial superconductivity.

cond-mat.supr-con↗

arXiv2018

Analog Circuit Applications based on Ambipolar Graphene/MoTe2 Vertical Transistors

The current integrated circuit (IC) technology based on conventional MOS-FET (metal-oxide-semiconductor field-effect transistor) is approaching the limit of miniaturization with increasing demand on energy. Several analog circuit applications based on graphene FETs have been demonstrated with less components comparing to the conventional technology. However, low on/off current ratio caused by the semimetal nature of graphene has severely hindered its practical applications. Here we report a graphene/MoTe2 van der Waals (vdW) vertical transistor with V-shaped ambipolar field effect transfer characteristics to overcome this challenge. Investigations on temperature dependence of transport properties reveal that gate tunable asymmetric barriers of the devices are account for the ambipolar behaviors. Furthermore, to demonstrate the analog circuit applications of such vdW vertical transistors, we successfully realized output polarity controllable (OPC) amplifier and frequency doubler. These results enable vdW heterojunction based electronic devices to open up new possibilities for wide perspective in telecommunication field.

cond-mat.mes-hall↗

arXiv2018

Robust memristors based on layered two-dimensional materials

Van der Waals heterostructure based on layered two-dimensional (2D) materials offers unprecedented opportunities to create materials with atomic precision by design. By combining superior properties of each component, such heterostructure also provides possible solutions to address various challenges of the electronic devices, especially those with vertical multilayered structures. Here, we report the realization of robust memristors for the first time based on van der Waals heterostructure of fully layered 2D materials (graphene/MoS2-xOx/graphene) and demonstrate a good thermal stability lacking in traditional memristors. Such devices have shown excellent switching performance with endurance up to 107 and a record-high operating temperature up to 340oC. By combining in situ high-resolution TEM and STEM studies, we have shown that the MoS2-xOx switching layer, together with the graphene electrodes and their atomically sharp interfaces, are responsible for the observed thermal stability at elevated temperatures. A well-defined conduction channel and a switching mechanism based on the migration of oxygen ions were also revealed. In addition, the fully layered 2D materials offer a good mechanical flexibility for flexible electronic applications, manifested by our experimental demonstration of a good endurance against over 1000 bending cycles. Our results showcase a general and encouraging pathway toward engineering desired device properties by using 2D van der Waals heterostructures.

cond-mat.mtrl-sci↗

arXiv2017

Nonparametric Variational Auto-encoders for Hierarchical Representation Learning

The recently developed variational autoencoders (VAEs) have proved to be an effective confluence of the rich representational power of neural networks with Bayesian methods. However, most work on VAEs use a rather simple prior over the latent variables such as standard normal distribution, thereby restricting its applications to relatively simple phenomena. In this work, we propose hierarchical nonparametric variational autoencoders, which combines tree-structured Bayesian nonparametric priors with VAEs, to enable infinite flexibility of the latent representation space. Both the neural parameters and Bayesian priors are learned jointly using tailored variational inference. The resulting model induces a hierarchical structure of latent semantic concepts underlying the data corpus, and infers accurate representations of data instances. We apply our model in video representation learning. Our method is able to discover highly interpretable activity hierarchies, and obtain improved clustering accuracy and generalization capacity based on the learned rich representations.

cs.LG↗

arXiv2016

Investigation of NWC-induced electron precipitation and theoretical simulation

Enhancement of the electron fluxes in the inner radiation belt, which is induced by the powerful North West Cape (NWC) very-low-frequency (VLF) transmitter, have been observed and analyzed by several research groups. However, all of the previous publications have focused on NWC-induced >100-keV electrons only, based on observations from the Detection of Electro-Magnetic Emissions Transmitted from Earthquake Regions (DEMETER) and the Geostationary Operational Environmental Satellite (GOES) satellites. Here, we present flux enhancements with 30--100-keV electrons related to NWC transmitter for the first time, which were observed by the GOES satellite at night. Similar to the 100--300-keV precipitated-electron havior, the low energy 30--100-keV electron precipitation is primarily located east of the transmitter. However, the latter does not drift eastward to the same extent as the former, possibly because of the lower electron velocity. The 30--100-keV electrons are distributed in the L=1.8--2.1 shell range, in contrast to the 100--300-keV electrons which are at L=1.67--1.9. This is consistent with the perspective that the energy of the VLF-wave-induced electron flux enhancement decreases with higher L-shell values. We expand upon the rationality of the simultaneous enhancement of the 30--100- and 100--300-keV electron fluxes through comparison with the cyclotron resonance theory for the quasi-linear wave-particle interaction. In addition, we interpret the asymmetry characteristics of NWC electric power distribution in north and south hemisphere by ray tracing model. Finally, we present considerable discussion and show that good agreement exists between the observation of satellites and theory.

physics.space-ph↗

arXiv2014

Study of typical space wave-particle coupling events possibly related with seismic activity

Based on the DEMETER satellite, we found two space wave-particle coupling events during February 2010 taking place in the range of McIlwain parameter $L$($1.27\sim1.37$). There are strong spatial and temporal correlation between the particle bursts(PBs) and the electromagnetic disturbances of the coupling events. The two PBs show different energy spectrum characteristics, while the corresponding electromagnetic disturbances concentrated on different frequencies range. In agreement with the prediction of the theory of wave-particle interaction, we conclude that the two wave-particle interactions can be probably explained as following: one is electron dominant precipitation with energy of $0.09\sim0.2$ MeV induced by VLF electromagnetic wave with the frequency of $14\sim20$ kHz, and another is proton dominant precipitation with energy of $0.65\sim2.85$ MeV induced by VLF electromagnetic wave with the frequency of $\le 100$ Hz. For the first time, those particle bursts origin, from electron or proton detected by the Instrument for the Detection of Particles(IDP) on board, is inferred by theory calculation, although the Instrument has no ability to identify the particle species.

physics.geo-ph↗