arXiv subjects

Ke Wang

Publications and source records attributed to Ke Wang.

At least 577 records · Page 32Linked to original sources

arXiv2020

Learning Semantic Program Embeddings with Graph Interval Neural Network

Learning distributed representations of source code has been a challenging task for machine learning models. Earlier works treated programs as text so that natural language methods can be readily applied. Unfortunately, such approaches do not capitalize on the rich structural information possessed by source code. Of late, Graph Neural Network (GNN) was proposed to learn embeddings of programs from their graph representations. Due to the homogeneous and expensive message-passing procedure, GNN can suffer from precision issues, especially when dealing with programs rendered into large graphs. In this paper, we present a new graph neural architecture, called Graph Interval Neural Network (GINN), to tackle the weaknesses of the existing GNN. Unlike the standard GNN, GINN generalizes from a curated graph representation obtained through an abstraction method designed to aid models to learn. In particular, GINN focuses exclusively on intervals for mining the feature representation of a program, furthermore, GINN operates on a hierarchy of intervals for scaling the learning to large graphs. We evaluate GINN for two popular downstream applications: variable misuse prediction and method name prediction. Results show in both cases GINN outperforms the state-of-the-art models by a comfortable margin. We have also created a neural bug detector based on GINN to catch null pointer deference bugs in Java code. While learning from the same 9,000 methods extracted from 64 projects, GINN-based bug detector significantly outperforms GNN-based bug detector on 13 unseen test projects. Next, we deploy our trained GINN-based bug detector and Facebook Infer to scan the codebase of 20 highly starred projects on GitHub. Through our manual inspection, we confirm 38 bugs out of 102 warnings raised by GINN-based bug detector compared to 34 bugs out of 129 warnings for Facebook Infer.

cs.SE↗

arXiv2020

Robust Visual Object Tracking with Two-Stream Residual Convolutional Networks

The current deep learning based visual tracking approaches have been very successful by learning the target classification and/or estimation model from a large amount of supervised training data in offline mode. However, most of them can still fail in tracking objects due to some more challenging issues such as dense distractor objects, confusing background, motion blurs, and so on. Inspired by the human "visual tracking" capability which leverages motion cues to distinguish the target from the background, we propose a Two-Stream Residual Convolutional Network (TS-RCN) for visual tracking, which successfully exploits both appearance and motion features for model update. Our TS-RCN can be integrated with existing deep learning based visual trackers. To further improve the tracking performance, we adopt a "wider" residual network ResNeXt as its feature extraction backbone. To the best of our knowledge, TS-RCN is the first end-to-end trainable two-stream visual tracking system, which makes full use of both appearance and motion features of the target. We have extensively evaluated the TS-RCN on most widely used benchmark datasets including VOT2018, VOT2019, and GOT-10K. The experiment results have successfully demonstrated that our two-stream model can greatly outperform the appearance based tracker, and it also achieves state-of-the-art performance. The tracking system can run at up to 38.1 FPS.

cs.CV↗

arXiv2020

Improving mobility of silicon metal-oxide-semiconductor devices for quantum dots by high vacuum activation annealing

To improve mobility of fabricated silicon metal-oxide-semiconductor (MOS) quantum devices, forming gas annealing is a common method used to mitigate the effects of disorder at the Si/SiO2 interface. However, the importance of activation annealing is usually ignored. Here, we show that a high vacuum environment for implantation activation is beneficial for improving mobility compared to nitrogen atmosphere. Low-temperature transport measurements of Hall bars show that peak mobility can be improved by a factor of two, reaching 1.5 m^2/(Vs) using high vacuum annealing during implantation activation. Moreover, the charge stability diagram of a single quantum dot is mapped, with no visible disturbance caused by disorder, suggesting possibility of fabricating high-quality quantum dots on commercial wafers. Our results may provide valuable insights into device optimization in silicon-based quantum computing.

cond-mat.mes-hall↗

arXiv2020

Hole spin in tunable Ge hut wire double quantum dot

Holes in germanium (Ge) exhibit strong spin-orbit interaction, which can be exploited for fast and all-electrical manipulation of spin states. Here, we report transport experiments in a tunable Ge hut wire hole double quantum dot. We observe the signatures of Pauli spin blockade (PSB) with a large singlet-triplet energy splitting of ~1.1 meV and extract the g factor. By analyzing the the PSB leakage current, we obtain a spin-orbit length l_so of ~ 40-100 nm. Furthermore, we demonstrate the electric dipole spin resonance. These results lay a solid foundation for implementing high quality tunable hole spin-orbit qubits.

cond-mat.mes-hall↗

arXiv2020

Tunable Ferromagnetism and Thermally Induced Spin Flip in Vanadium-doped Tungsten Diselenide Monolayers at Room Temperature

The outstanding optoelectronic and valleytronic properties of transition metal dichalcogenides (TMDs) have triggered intense research efforts by the scientific community. An alternative to induce long-range ferromagnetism (FM) in TMDs is by introducing magnetic dopants to form a dilute magnetic semiconductor. Enhancing ferromagnetism in these semiconductors not only represents a key step towards modern TMD-based spintronics, but also enables exploration of new and exciting dimensionality-driven magnetic phenomena. To this end, we show tunable ferromagnetism at room temperature and a thermally induced spin flip (TISF) in monolayers of V-doped WSe2. As vanadium concentrations increase within the WSe2 monolayers the saturation magnetization increases, and it is optimal at ~4at.% vanadium; the highest doping/alloying level ever achieved for V-doped WSe2 monolayers. The TISF occurs at ~175 K and becomes more pronounced upon increasing the temperature towards room temperature. We demonstrate that TISF can be manipulated by changing the vanadium concentration within the WSe2 monolayers. We attribute TISF to the magnetic field and temperature dependent flipping of the nearest W-site magnetic moments that are antiferromagnetically coupled to the V magnetic moments in the ground state. This is fully supported by a recent spin-polarized density functional theory calculation. Our findings pave the way for the development of novel spintronic and valleytronic nanodevices based on atomically thin magnetic semiconductors and stimulate further studies in this rapidly expanding research field of 2D magnetism.

cond-mat.mtrl-sci↗

arXiv2020

ALMA ACA and Nobeyama observations of two Orion cores in deuterated molecular lines

We mapped two molecular cloud cores in the Orion A cloud with the ALMA ACA 7-m Array and with the Nobeyama 45-m radio telescope. These cores have bright N$_2$D$^+$ emission in single-pointing observations with the Nobeyama 45-m radio telescope, have relatively high deuterium fraction, and are thought to be close to the onset of star formation. One is a star-forming core, and the other is starless. These cores are located along filaments observed in N$_2$H$^+$, and show narrow linewidths of 0.41 km s$^{-1}$ and 0.45 km s$^{-1}$ in N$_2$D$^+$, respectively, with the Nobeyama 45-m telescope. Both cores were detected with the ALMA ACA 7m Array in the continuum and molecular lines at Band 6. The starless core G211 shows clumpy structure with several sub-cores, which in turn show chemical differences. Also, the sub-cores in G211 have internal motions that are almost purely thermal. The starless sub-core G211D, in particular, shows a hint of the inverse P Cygni profile, suggesting infall motion. The star-forming core G210 shows an interesting spatial feature of two N$_2$D$^+$ peaks of similar intensity and radial velocity located symmetrically with respect to the single dust continuum peak. One interpretation is that the two N$_2$D$^+$ peaks represent an edge-on pseudo-disk. The CO outflow lobes, however, are not directed perpendicular to the line connecting both N$_2$D$^+$ peaks.

astro-ph.GA↗

arXiv2020

Edge collapse and subsequent longitudinal accretion in Filament S242

Filament S242 is 25 pc long with massive clumps and YSO clusters concentrated in its end regions; it is considered a good example of edge collapse. We mapped this filament in the $^{12}$CO(1-0) and $^{13}$CO(1-0) lines. A large-scale velocity gradient along filament S242 has been detected; the relative velocity between the two end-clumps is $\sim$ 3 km s$^{-1}$, indicating an approaching motion between them. These signatures are consistent with the filament S242 being formed through the collapse of a single elongated entity, where an effect known as "gravitational focusing" drives the ends of the filament to collapse (edge collapse). Based on this picture, we estimate a collapse timescale of $\sim$ 4.2 Myr, which is the time needed for a finite and elongated entity evolving to the observed filament S242. For the whole filament, we find that increases in surface densities lead to increases in velocity dispersion, which can be consistently explained as the result of self-gravity. We also calculated the contribution of longitudinal collapse to the observed velocity dispersion and found it to be the dominant effect in driving the gas motion near the end-clumps. We propose that our filament S242 is formed through a two-stage collapse model, where the edge collapse of a truncated filament is followed by a stage of longitudinal accretion toward the dense end-clumps.

astro-ph.GA↗

arXiv2020

Unexpected two-fold symmetric superconductivity in few-layer NbSe$_2$

Two-dimensional transition metal dichalcogenides (TMDs) have been attracting significant interest due to a range of properties, such as layer-dependent inversion symmetry, valley-contrasted Berry curvatures, and strong spin-orbit coupling (SOC). Of particular interest is niobium diselenide (NbSe2), whose superconducting state in few-layer samples is profoundly affected by an unusual type of SOC called Ising SOC. Combined with the reduced dimensionality, the latter stabilizes the superconducting state against magnetic fields up to ~35 T and could lead to other exotic properties such as nodal and crystalline topological superconductivity. Here, we report transport measurements of few-layer NbSe$_2$ under in-plane external magnetic fields, revealing an unexpected two-fold rotational symmetry of the superconducting state. In contrast to the three-fold symmetry of the lattice, we observe that the magnetoresistance and critical field exhibit a two-fold oscillation with respect to an applied in-plane magnetic field. We find similar two-fold oscillations deep inside the superconducting state in differential conductance measurements on NbSe$_2$/CrBr$_3$ superconductor-magnet junctions. In both cases, the anisotropy vanishes in the normal state, demonstrating that it is an intrinsic property of the superconducting phase. We attribute the behavior to the mixing between two closely competing pairing instabilities, namely, the conventional s-wave instability typical of bulk NbSe$_2$ and an unconventional d- or p-wave channel that emerges in few-layer NbSe2. Our results thus demonstrate the unconventional character of the pairing interaction in a few-layer TMD, opening a new avenue to search for exotic superconductivity in this family of 2D materials.

cond-mat.supr-con↗

arXiv2020

Boundary feedback stabilization of quasilinear hyperbolic systems with partially dissipative structure

In this paper, we study the boundary feedback stabilization of a quasilinear hyperbolic system with partially dissipative structure. Thanks to this structure, we construct a suitable Lyapunov function which leads to the exponential stability to the equilibrium of the $H^2$ solution. As an application, we also obtain the feedback stabilization for the Saint-Venant-Exner model under physical boundary conditions.

math.OC↗

arXiv2020

Learning a Static Bug Finder from Data

We present an alternative approach to creating static bug finders. Instead of relying on human expertise, we utilize deep neural networks to train static analyzers directly from data. In particular, we frame the problem of bug finding as a classification task and train a classifier to differentiate the buggy from non-buggy programs using Graph Neural Network (GNN). Crucially, we propose a novel interval-based propagation mechanism that leads to a significantly more efficient, accurate and scalable generalization of GNN. We have realized our approach into a framework, NeurSA, and extensively evaluated it. In a cross-project prediction task, three neural bug detectors we instantiate from NeurSA are effective in catching null pointer dereference, array index out of bound and class cast bugs in unseen code. We compare NeurSA against several static analyzers (e.g. Facebook Infer and Pinpoint) on a set of null pointer dereference bugs. Results show that NeurSA is more precise in catching the real bugs and suppressing the spurious warnings. We also apply NeurSA to several popular Java projects on GitHub and discover 50 new bugs, among which 9 have been fixed, and 3 have been confirmed.

cs.SE↗

arXiv2020

CF2-Net: Coarse-to-Fine Fusion Convolutional Network for Breast Ultrasound Image Segmentation

Breast ultrasound (BUS) image segmentation plays a crucial role in a computer-aided diagnosis system, which is regarded as a useful tool to help increase the accuracy of breast cancer diagnosis. Recently, many deep learning methods have been developed for segmentation of BUS image and show some advantages compared with conventional region-, model-, and traditional learning-based methods. However, previous deep learning methods typically use skip-connection to concatenate the encoder and decoder, which might not make full fusion of coarse-to-fine features from encoder and decoder. Since the structure and edge of lesion in BUS image are common blurred, these would make it difficult to learn the discriminant information of structure and edge, and reduce the performance. To this end, we propose and evaluate a coarse-to-fine fusion convolutional network (CF2-Net) based on a novel feature integration strategy (forming an 'E'-like type) for BUS image segmentation. To enhance contour and provide structural information, we concatenate a super-pixel image and the original image as the input of CF2-Net. Meanwhile, to highlight the differences in the lesion regions with variable sizes and relieve the imbalance issue, we further design a weighted-balanced loss function to train the CF2-Net effectively. The proposed CF2-Net was evaluated on an open dataset by using four-fold cross validation. The results of the experiment demonstrate that the CF2-Net obtains state-of-the-art performance when compared with other deep learning-based methods

eess.IV↗

arXiv2020

FPGA-based Neural Network Accelerator for Millimeter-Wave Radio-over-Fiber Systems

With the rapidly-developing high-speed wireless communications, the 60 GHz millimeter-wave frequency range and radio-over-fiber systems have been investigated as a promising solution to deliver mm-wave signals. Neural networks have been studied to improve the mm-wave RoF system performances at the receiver side by suppressing linear and nonlinear impairments. However, previous neural network studies in mm-wave RoF systems focus on the off-line implementation with high-end GPUs , which is not practical for low power-consumption, low-cost and limited computation platform applications. To solve this issue, we investigate neural network hardware accelerator implementations using the field programmable gate array (FPGA), taking advantage of the low power consumption, parallel computation capability, and reconfigurablity features of FPGA. Convolutional neural network (CNN) and binary convolutional neural network (BCNN) hardware accelerators are demonstrated. In addition, to satisfy the low-latency requirement in mm-wave RoF systems and to enable the use of low-cost compact FPGA devices, a novel inner parallel optimization method is proposed. Compared with the embedded processor (ARM Cortex A9) execution latency, the CNN/BCNN FPGA-based hardware accelerator reduces their latency by over 92%. Compared with non-optimized FPGA implementations, the proposed optimization method reduces the processing latency by over 44% for CNN and BCNN. Compared with the GPU implementation, the latency of CNN implementation with the proposed optimization method is reduced by 85.49%, while the power consumption is reduced by 86.91%. Although the latency of BCNN implementation with the proposed optimization method is larger compared with the GPU implementation, the power consumption is reduced by 86.14%. The FPGA-based neural network hardware accelerators provide a promising solution for mm-wave RoF systems.

eess.SP↗

arXiv2020

The Differentially Private Lottery Ticket Mechanism

We propose the differentially private lottery ticket mechanism (DPLTM). An end-to-end differentially private training paradigm based on the lottery ticket hypothesis. Using "high-quality winners", selected via our custom score function, DPLTM significantly improves the privacy-utility trade-off over the state-of-the-art. We show that DPLTM converges faster, allowing for early stopping with reduced privacy budget consumption. We further show that the tickets from DPLTM are transferable across datasets, domains, and architectures. Our extensive evaluation on several public datasets provides evidence to our claims.

cs.LG↗

arXiv2020

SEED RL: Scalable and Efficient Deep-RL with Accelerated Central Inference

We present a modern scalable reinforcement learning agent called SEED (Scalable, Efficient Deep-RL). By effectively utilizing modern accelerators, we show that it is not only possible to train on millions of frames per second but also to lower the cost of experiments compared to current methods. We achieve this with a simple architecture that features centralized inference and an optimized communication layer. SEED adopts two state of the art distributed algorithms, IMPALA/V-trace (policy gradients) and R2D2 (Q-learning), and is evaluated on Atari-57, DeepMind Lab and Google Research Football. We improve the state of the art on Football and are able to reach state of the art on Atari-57 three times faster in wall-time. For the scenarios we consider, a 40% to 80% cost reduction for running experiments is achieved. The implementation along with experiments is open-sourced so results can be reproduced and novel ideas tried out.

cs.LG↗

arXiv2020

Differentially Private Survival Function Estimation

Survival function estimation is used in many disciplines, but it is most common in medical analytics in the form of the Kaplan-Meier estimator. Sensitive data (patient records) is used in the estimation without any explicit control on the information leakage, which is a significant privacy concern. We propose a first differentially private estimator of the survival function and show that it can be easily extended to provide differentially private confidence intervals and test statistics without spending any extra privacy budget. We further provide extensions for differentially private estimation of the competing risk cumulative incidence function, Nelson-Aalen's estimator for the hazard function, etc. Using eleven real-life clinical datasets, we provide empirical evidence that our proposed method provides good utility while simultaneously providing strong privacy guarantees.

cs.LG↗

arXiv2020

Experimental Demonstration of Millimeter-Wave Radio-over-Fiber System with Convolutional Neural Network (CNN) and Binary Convolutional Neural Network (BCNN)

The millimeter-wave (mm-wave) radio-over-fiber (RoF) systems have been widely studied as promising solutions to deliver high-speed wireless signals to end users, and neural networks have been studied to solve various linear and nonlinear impairments. However, high computation cost and large amounts of training data are required to effectively improve the system performance. In this paper, we propose and demonstrate highly computation efficient convolutional neural network (CNN) and binary convolutional neural network (BCNN) based decision schemes to solve these limitations. The proposed CNN and BCNN based decision schemes are demonstrated in a 5 Gbps 60 GHz RoF system for up to 20 km fiber distance. Compared with previously demonstrated neural networks, results show that the bit error rate (BER) performance and the computation intensive training process are improved. The number of training iterations needed is reduced by about 50 % and the amount of required training data is reduced by over 30 %. In addition, only one training is required for the entire measured received optical power range over 3.5 dB in the proposed CNN and BCNN schemes, to further reduce the computation cost of implementing neural networks decision schemes in mm-wave RoF systems.

eess.SP↗

arXiv2020

Concurrence of Quantum Anomalous Hall and Topological Hall Effects in Magnetic Topological Insulator Sandwich Heterostructures

The quantum anomalous Hall (QAH) effect is a quintessential consequence of non-zero Berry curvature in momentum-space. The QAH insulator harbors dissipation-free chiral edge states in the absence of an external magnetic field. On the other hand, the topological Hall (TH) effect, a transport hallmark of the chiral spin textures, is a consequence of real-space Berry curvature. While both the QAH and TH effects have been reported separately, their coexistence, a manifestation of entangled chiral edge states and chiral spin textures, has not been reported. Here, by inserting a TI layer between two magnetic TI layers to form a sandwich heterostructure, we realized a concurrence of the TH effect and the QAH effect through electric field gating. The TH effect is probed by bulk carriers, while the QAH effect is characterized by chiral edge states. The appearance of TH effect in the QAH insulating regime is the consequence of chiral magnetic domain walls that result from the gate-induced Dzyaloshinskii-Moriya interaction and occur during the magnetization reversal process in the magnetic TI sandwich samples. The coexistence of chiral edge states and chiral spin textures potentially provides a unique platform for proof-of-concept dissipationless spin-textured spintronic applications.

cond-mat.mes-hall↗

arXiv2019

ALMA observations reveal no preferred outflow--filament and outflow--magnetic field orientations

We present a statistical study on the orientation of outflows with respect to large-scale filaments and the magnetic fields. Although filaments are widely observed toward Galactic star-forming regions, the exact role of filaments in star formation is unclear. Studies toward low-mass star-forming regions revealed both preferred and random orientation of outflows respective to the filament long-axes, while outflows in massive star-forming regions mostly oriented perpendicular to the host filaments, and parallel to the magnetic fields at similar physical scales. Here, we explore outflows in a sample of 11 protoclusters in HII regions, a more evolved stage compared to IRDCs, using ALMA CO (3-2) line observations. We identify a total of 105 outflow lobes in these protoclusters. Among the 11 targets, 7 are embedded within parsec-scale filamentary structures detected in $^{13}$CO line and 870 $μm$ continuum emissions. The angles between outflow axes and corresponding filaments ($γ_\mathrm{Fil}$) do not show any hint of preferred orientations (i.e., orthogonal or parallel as inferred in numerical models) with respect to the position angle of the filaments. Identified outflow lobes are also not correlated with the magnetic fields and Galactic plane position angles. Outflows associated with filaments aligned along the large-scale magnetic fields are also randomly orientated. Our study presents the first statistical results of outflow orientation respective to large-scale filaments and magnetic fields in evolved massive star-forming regions. The random distribution suggests a lack of alignment of outflows with filaments, which may be a result of the evolutionary stage of the clusters.

astro-ph.GA↗