arXiv subjects

Peng Li

Publications and source records attributed to Peng Li.

At least 451 records · Page 25Linked to original sources

arXiv2019

Broadband Optical Detection using the Spin Seebeck Effect

The generation, control, and detection of spin currents in solid-state devices are critical for Joule-heating minimization, spin-based computation, and electrical energy generation from thermal gradients. Although incorporation of spin functionality into technologically important architectures is still in its infancy, advantages over all-electric devices are increasingly becoming clear. Here, we utilize the spin Seebeck effect (SSE) in Pt/Y3Fe5O12 devices to detect light from 390 to 2200 nm. We find the device responsivity is remarkably flat across this technologically important wavelength range, closely following the Pt absorption coefficient. As expected from a SSE-generation mechanism, we observe that the photovoltage and Pt heating dynamics are in strong agreement. To precisely determine the optically created thermal gradient produced from a point-like heat source, we introduce a field-modulation method for measuring the SSE. Our results show broadband optical detection can be performed with devices based solely on spin current generation and detection.

cond-mat.mtrl-sci↗

arXiv2019

FastPose: Towards Real-time Pose Estimation and Tracking via Scale-normalized Multi-task Networks

Both accuracy and efficiency are significant for pose estimation and tracking in videos. State-of-the-art performance is dominated by two-stages top-down methods. Despite the leading results, these methods are impractical for real-world applications due to their separated architectures and complicated calculation. This paper addresses the task of articulated multi-person pose estimation and tracking towards real-time speed. An end-to-end multi-task network (MTN) is designed to perform human detection, pose estimation, and person re-identification (Re-ID) tasks simultaneously. To alleviate the performance bottleneck caused by scale variation problem, a paradigm which exploits scale-normalized image and feature pyramids (SIFP) is proposed to boost both performance and speed. Given the results of MTN, we adopt an occlusion-aware Re-ID feature strategy in the pose tracking module, where pose information is utilized to infer the occlusion state to make better use of Re-ID feature. In experiments, we demonstrate that the pose estimation and tracking performance improves steadily utilizing SIFP through different backbones. Using ResNet-18 and ResNet-50 as backbones, the overall pose tracking framework achieves competitive performance with 29.4 FPS and 12.2 FPS, respectively. Additionally, occlusion-aware Re-ID feature decreases the identification switches by 37% in the pose tracking process.

cs.CV↗

arXiv2019

General approaches for shear-correcting coordinate transformations in Bragg coherent diffraction imaging: Part 1

In this two-part article series we provide a generalized description of the scattering geometry of Bragg coherent diffraction imaging (BCDI) experiments, the shear distortion effects inherent to the resulting three-dimensional (3D) image from current phase retrieval methods and strategies to mitigate this distortion. In this Part I, we derive in general terms the real-space coordinate transformation to correct this shear, which originates in the more fundamental relationship between the representations of mutually conjugate 3D spaces. Such a transformation, applied as a final post-processing step following phase retrieval, is crucial for arriving at an un-distorted and physically meaningful image of the 3D scatterer. As the relevance of BCDI grows in the field of materials characterization, we take this opportunity to generalize the available sparse literature that addresses the geometric theory of BCDI and the subsequent analysis methods. This aspect, specific to coherent Bragg diffraction and absent in two-dimensional transmission CDI experiments, gains particular importance concerning spatially-resolved characterization of 3D crystalline materials in a realiable, non-destructive manner. These articles describe this theory, from the diffraction in Bragg geometry, to the corrections needed to obtain a properly rendered digital image of the 3D scatterer. Part I provides the experimental BCDI communitcy with the theoretical underpinnings of the 3D real-space distortions in the phase-retrieved object, along with the necessary post-retrieval correction method. Part II builds upon the geometric theory developed in Part I with the formalism to correct the shear distortions directly on an orthogonal grid within the phase retrieval algorithm itself, allowing more physically realistic constraints to be applied.

physics.ins-det↗

arXiv2019

General approaches for shear-correcting coordinate transformations in Bragg coherent diffraction imaging: Part 2

X-ray Bragg coherent diffraction imaging has been demonstrated as a powerful three-dimensional (3D) microscopy approach for the investigation of sub-micrometer-scale crystalline particles. It is based on the measurement of a series of coherent diffraction intensity patterns that are numerically inverted to retrieve an image of the spatial distribution of relative phase and amplitude of the Bragg structure factor of the scatterer. This 3D information, which is collected through an angular rotation of the sample, is necessarily obtained in a non-orthogonal frame in Fourier space that must be eventually reconciled. To deal with this, the currently favored approach (detailed in Part I) is to perform the entire inversion in conjugate non-orthogonal real and Fourier space frames, and to transform the 3D sample image into an orthogonal frame as a post-processing step for result analysis. In this article, a direct follow-up of Part I, we demonstrate two different transformation strategies that enable the entire inversion procedure of the measured data set to be performed in an orthogonal frame. The new approaches described here build mathematical and numerical frameworks that apply to the cases of evenly and non-evenly sampled data along the direction of sample rotation (the rocking curve). The value of these methods is that they rely on and incorporate significantly more information about the experimental geometry into the design of the phase retrieval Fourier transformation than the strategy presented in Part I. Two important outcomes are 1) that the resulting sample image is correctly interpreted in a shear-free frame, and 2) physically realistic constraints of BCDI phase retrieval that are difficult to implement with current methods are easily incorporated. Computing scripts are also given to aid readers in the implementation of the proposed formalisms.

physics.ins-det↗

arXiv2019

DocRED: A Large-Scale Document-Level Relation Extraction Dataset

Multiple entities in a document generally exhibit complex inter-sentence relations, and cannot be well handled by existing relation extraction (RE) methods that typically focus on extracting intra-sentence relations for single entity pairs. In order to accelerate the research on document-level RE, we introduce DocRED, a new dataset constructed from Wikipedia and Wikidata with three features: (1) DocRED annotates both named entities and relations, and is the largest human-annotated dataset for document-level RE from plain text; (2) DocRED requires reading multiple sentences in a document to extract entities and infer their relations by synthesizing all information of the document; (3) along with the human-annotated data, we also offer large-scale distantly supervised data, which enables DocRED to be adopted for both supervised and weakly supervised scenarios. In order to verify the challenges of document-level RE, we implement recent state-of-the-art methods for RE and conduct a thorough evaluation of these methods on DocRED. Empirical results show that DocRED is challenging for existing RE methods, which indicates that document-level RE remains an open problem and requires further efforts. Based on the detailed analysis on the experiments, we discuss multiple promising directions for future research.

cs.CL↗

arXiv2019

Key Fact as Pivot: A Two-Stage Model for Low Resource Table-to-Text Generation

Table-to-text generation aims to translate the structured data into the unstructured text. Most existing methods adopt the encoder-decoder framework to learn the transformation, which requires large-scale training samples. However, the lack of large parallel data is a major practical problem for many domains. In this work, we consider the scenario of low resource table-to-text generation, where only limited parallel data is available. We propose a novel model to separate the generation into two stages: key fact prediction and surface realization. It first predicts the key facts from the tables, and then generates the text with the key facts. The training of key fact prediction needs much fewer annotated data, while surface realization can be trained with pseudo parallel corpus. We evaluate our model on a biography generation dataset. Our model can achieve $27.34$ BLEU score with only $1,000$ parallel data, while the baseline model only obtain the performance of $9.71$ BLEU score.

cs.CL↗

arXiv2019

Strong band kinks in magic-thickness Yb films arising from interfacial electron-phonon coupling

Interfacial electron-phonon coupling in ultrathin films has attracted much interest; it can give rise to novel effects and phenomena, including enhanced superconductivity. Here we report an observation of strong kinks in the energy dispersions of quantum well states in ultrathin Yb films grown on graphite. These kinks, arising from interfacial electron-phonon coupling, are most prominent for films with a preferred ("magic") thickness of 4 monolayers, which are strained and hole doped by the substrate. The energy position of the kinks agrees well with the optical phonon energy of graphite, and the extracted electron-phonon coupling strength λ shows a large subband dependence, with a maximum value up to 0.6. The kinks decay rapidly with increasing film thickness, consistent with its interfacial origin. The variation of λ is correlated with evolution of the electronic wave function amplitudes at the interface. A Lifshitz transition occurs just beyond the magic thickness where the heavy Yb 5d bands begin to populate right below the Fermi level.

cond-mat.mes-hall↗

arXiv2019

Global Adversarial Attacks for Assessing Deep Learning Robustness

It has been shown that deep neural networks (DNNs) may be vulnerable to adversarial attacks, raising the concern on their robustness particularly for safety-critical applications. Recognizing the local nature and limitations of existing adversarial attacks, we present a new type of global adversarial attacks for assessing global DNN robustness. More specifically, we propose a novel concept of global adversarial example pairs in which each pair of two examples are close to each other but have different class labels predicted by the DNN. We further propose two families of global attack methods and show that our methods are able to generate diverse and intriguing adversarial example pairs at locations far from the training or testing data. Moreover, we demonstrate that DNNs hardened using the strong projected gradient descent (PGD) based (local) adversarial training are vulnerable to the proposed global adversarial example pairs, suggesting that global robustness must be considered while training robust deep learning networks.

cs.LG↗

arXiv2019

SAVIOR: Towards Bug-Driven Hybrid Testing

Hybrid testing combines fuzz testing and concolic execution. It leverages fuzz testing to test easy-to-reach code regions and uses concolic execution to explore code blocks guarded by complex branch conditions. However, its code coverage-centric design is inefficient in vulnerability detection. First, it blindly selects seeds for concolic execution and aims to explore new code continuously. However, as statistics show, a large portion of the explored code is often bug-free. Therefore, giving equal attention to every part of the code during hybrid testing is a non-optimal strategy. It slows down the detection of real vulnerabilities by over 43%. Second, classic hybrid testing quickly moves on after reaching a chunk of code, rather than examining the hidden defects inside. It may frequently miss subtle vulnerabilities despite that it has already explored the vulnerable code paths. We propose SAVIOR, a new hybrid testing framework pioneering a bug-driven principle. Unlike the existing hybrid testing tools, SAVIOR prioritizes the concolic execution of the seeds that are likely to uncover more vulnerabilities. Moreover, SAVIOR verifies all vulnerable program locations along the executing program path. By modeling faulty situations using SMT constraints, SAVIOR reasons the feasibility of vulnerabilities and generates concrete test cases as proofs. Our evaluation shows that the bug-driven approach outperforms mainstream automated testing techniques, including state-of-the-art hybrid testing systems driven by code coverage. On average, SAVIOR detects vulnerabilities 43.4% faster than DRILLER and 44.3% faster than QSYM, leading to the discovery of 88 and 76 more uniquebugs,respectively.Accordingtotheevaluationon11 well fuzzed benchmark programs, within the first 24 hours, SAVIOR triggers 481 UBSAN violations, among which 243 are real bugs.

cs.SE↗

arXiv2019

Large Fermi Surface Expansion through Anisotropic c-f Mixing in the Semimetallic Kondo Lattice System CeBi

Using angle-resolved photoemission spectroscopy (ARPES) and resonant ARPES, we report evidence of strong anisotropic conduction-f electron mixing (c-f mixing) in CeBi by observing a largely expanded Ce-5d pocket at low temperature, with no change in the Bi-6p bands. The Fermi surface (FS) expansion is accompanied by a pronounced spectral weight transfer from the local 4f 0 peak of Ce (corresponding to Ce3+) to the itinerant conduction bands near the Fermi level. Careful analysis suggests that the observed large FS change (with a volume expansion of the electron pocket up to 40%) can most naturally be explained by a small valence change (~ 1%) of Ce, which coexists with a very weak Kondo screening. Our work therefore provides evidence for a FS change driven by real charge fluctuations deep in the Kondo limit, which is made possible by the low carrier density.

cond-mat.str-el↗

arXiv2019

Exploiting Offset-guided Network for Pose Estimation and Tracking

Human pose estimation has witnessed a significant advance thanks to the development of deep learning. Recent human pose estimation approaches tend to directly predict the location heatmaps, which causes quantization errors and inevitably deteriorates the performance within the reduced network output. Aim at solving it, we revisit the heatmap-offset aggregation method and propose the Offset-guided Network (OGN) with an intuitive but effective fusion strategy for both two-stages pose estimation and Mask R-CNN. For two-stages pose estimation, a greedy box generation strategy is also proposed to keep more necessary candidates while performing person detection. For mask R-CNN, ratio-consistent is adopted to improve the generalization ability of the network. State-of-the-art results on COCO and PoseTrack dataset verify the effectiveness of our offset-guided pose estimation and tracking.

cs.CV↗

arXiv2019

State-aware Re-identification Feature for Multi-target Multi-camera Tracking

Multi-target Multi-camera Tracking (MTMCT) aims to extract the trajectories from videos captured by a set of cameras. Recently, the tracking performance of MTMCT is significantly enhanced with the employment of re-identification (Re-ID) model. However, the appearance feature usually becomes unreliable due to the occlusion and orientation variance of the targets. Directly applying Re-ID model in MTMCT will encounter the problem of identity switches (IDS) and tracklet fragment caused by occlusion. To solve these problems, we propose a novel tracking framework in this paper. In this framework, the occlusion status and orientation information are utilized in Re-ID model with human pose information considered. In addition, the tracklet association using the proposed fused tracking feature is adopted to handle the fragment problem. The proposed tracker achieves 81.3\% IDF1 on the multiple-camera hard sequence, which outperforms all other reference methods by a large margin.

cs.CV↗

arXiv2019

A Dual Reinforcement Learning Framework for Unsupervised Text Style Transfer

Unsupervised text style transfer aims to transfer the underlying style of text but keep its main content unchanged without parallel data. Most existing methods typically follow two steps: first separating the content from the original style, and then fusing the content with the desired style. However, the separation in the first step is challenging because the content and style interact in subtle ways in natural language. Therefore, in this paper, we propose a dual reinforcement learning framework to directly transfer the style of the text via a one-step mapping model, without any separation of content and style. Specifically, we consider the learning of the source-to-target and target-to-source mappings as a dual task, and two rewards are designed based on such a dual structure to reflect the style accuracy and content preservation, respectively. In this way, the two one-step mapping models can be trained via reinforcement learning, without any use of parallel data. Automatic evaluations show that our model outperforms the state-of-the-art systems by a large margin, especially with more than 8 BLEU points improvement averaged on two benchmark datasets. Human evaluations also validate the effectiveness of our model in terms of style accuracy, content preservation and fluency. Our code and data, including outputs of all baselines and our model are available at https://github.com/luofuli/DualLanST.

cs.CL↗

arXiv2019

SNR-adaptive OCT angiography enabled by statistical characterization of intensity and decorrelation with multi-variate time series model

In OCT angiography (OCTA), decorrelation computation has been widely used as a local motion index to identify dynamic flow from static tissues, but its dependence on SNR severely degrades the vascular visibility, particularly in low- SNR regions. To mathematically characterize the decorrelation-SNR dependence of OCT signals, we developed a multi-variate time series (MVTS) model. Based on the model, we derived a universal asymptotic linear relation of decorrelation to inverse SNR (iSNR), with the variance in static and noise regions determined by the average kernel size. Accordingly, with the population distribution of static and noise voxels being explicitly calculated in the iSNR and decorrelation (ID) space, a linear classifier is developed by removing static and noise voxels at all SNR, to generate a SNR-adaptive OCTA, termed as ID-OCTA. Then, flow phantom and human skin experiments were performed to validate the proposed ID-OCTA. Both qualitative and quantitative assessments demonstrated that ID-OCTA offers a superior visibility of blood vessels, particularly in the deep layer. Finally, implications of this work on both system design and hemodynamic quantification are further discussed.

physics.med-ph↗

arXiv2019

Development of an Autonomous Sanding Robot with Structured-Light Technology

Large demand for robotics and automation has been reflected in the sanding works, as current manual operations are labor-intensive, without consistent quality, and also subject to safety and health issues. While several machines have been developed to automate one or two steps in the sanding works, the autonomous capability of existing solutions is relatively low, and the human assistance or supervision is still heavily required in the calibration of target objects or the planning of robot motion and tasks. This paper presents the development of an autonomous sanding robot, which is able to perform the sanding works on an unknown object automatically, without any prior calibration or human intervention. The developed robot works as follows. First, the target object is scanned then modeled with the structured-light camera. Second, the robot motion is planned to cover all the surfaces of the object with an optimized transition sequence. Third, the robot is controlled to perform the sanding on the object under the desired impedance model. A prototype of the sanding robot is fabricated and its performance is validated in the task of sanding a batch of wooden boxes. With sufficient degrees of freedom (DOFs) and the module design for the end effector, the developed robot is able to provide a general solution to the autonomous sanding on many other different objects.

eess.SY↗

arXiv2019

Option Comparison Network for Multiple-choice Reading Comprehension

Multiple-choice reading comprehension (MCRC) is the task of selecting the correct answer from multiple options given a question and an article. Existing MCRC models typically either read each option independently or compute a fixed-length representation for each option before comparing them. However, humans typically compare the options at multiple-granularity level before reading the article in detail to make reasoning more efficient. Mimicking humans, we propose an option comparison network (OCN) for MCRC which compares options at word-level to better identify their correlations to help reasoning. Specially, each option is encoded into a vector sequence using a skimmer to retain fine-grained information as much as possible. An attention mechanism is leveraged to compare these sequences vector-by-vector to identify more subtle correlations between options, which is potentially valuable for reasoning. Experimental results on the human English exam MCRC dataset RACE show that our model outperforms existing methods significantly. Moreover, it is also the first model that surpasses Amazon Mechanical Turker performance on the whole dataset.

cs.CL↗

arXiv2019

Distributed and Application-aware Task Scheduling in Edge-clouds

Edge computing is an emerging technology which places computing at the edge of the network to provide an ultra-low latency. Computation offloading, a paradigm that migrates computing from mobile devices to remote servers, can now use the power of edge computing by offloading computation to cloudlets in edge-clouds. However, the task scheduling of computation offloading in edge-clouds faces a two-fold challenge. First, as cloudlets are geographically distributed, it is difficult for each cloudlet to perform load balancing without centralized control. Second, as tasks of computation offloading have a wide variety of types, to guarantee the user quality of experience (QoE) in terms of task types is challenging. In this paper, we present Petrel, a distributed and application-aware task scheduling framework for edge-clouds. Petrel implements a sample-based load balancing technology and further adopts adaptive scheduling policies according to task types. This application-aware scheduling not only provides QoE guarantee but also improves the overall scheduling performance. Trace-driven simulations show that Petrel achieves a significant improvement over existing scheduling strategies.

cs.DC↗

arXiv2019

Tunable Electronic Structure and Surface States in Rare Earth Mono-Bismuthides with Partially Filled f Shell

Here we report the evolution of bulk band structure and surface states in rare earth mono-bismuthides with partially filled f shell. Utilizing synchrotron-based photoemission spectroscopy, we determined the three-dimensional bulk band structure and identified the bulk band inversions near the X points, which, according to the topological theory, could give rise to nontrivial band topology with odd number of gapless topological surface states. Near the surface Gamma bar point, no clear evidence for predicted gapless topological surface state is observed due to its strong hybridization with the bulk bands. Near the M bar point, the two surface states, due to projections from two inequivalent bulk band inversions, interact and give rise to two peculiar sets of gapped surface states. The bulk band inversions and corresponding surface states can be tuned substantially by varying rare earth elements, in good agreement with density-functional theory calculations assuming local f electrons. Our study therefore establishes rare earth mono-bismuthides as an interesting class of materials possessing tunable electronic properties and magnetism, providing a promising platform to search for novel properties in potentially correlated topological materials.

cond-mat.mes-hall↗