arXiv subjects

Hao Wu

Publications and source records attributed to Hao Wu.

At least 757 records · Page 42Linked to original sources

arXiv2020

Response to LiveBot: Generating Live Video Comments Based on Visual and Textual Contexts

Live video commenting systems are an emerging feature of online video sites. Recently the Chinese video sharing platform Bilibili, has popularised a novel captioning system where user comments are displayed as streams of moving subtitles overlaid on the video playback screen and broadcast to all viewers in real-time. LiveBot was recently introduced as a novel Automatic Live Video Commenting (ALVC) application. This enables the automatic generation of live video comments from both the existing video stream and existing viewers comments. In seeking to reproduce the baseline results reported in the original Livebot paper, we found differences between the reproduced results using the project codebase and the numbers reported in the paper. Further examination of this situation suggests that this may be caused by a number of small issues in the project code, including a non-obvious overlap between the training and test sets. In this paper, we study these discrepancies in detail and propose an alternative baseline implementation as a reference for other researchers in this field.

cs.CL↗

arXiv2020

Modeling nanoconfinement effects using active learning

Predicting the spatial configuration of gas molecules in nanopores of shale formations is crucial for fluid flow forecasting and hydrocarbon reserves estimation. The key challenge in these tight formations is that the majority of the pore sizes are less than 50 nm. At this scale, the fluid properties are affected by nanoconfinement effects due to the increased fluid-solid interactions. For instance, gas adsorption to the pore walls could account for up to 85% of the total hydrocarbon volume in a tight reservoir. Although there are analytical solutions that describe this phenomenon for simple geometries, they are not suitable for describing realistic pores, where surface roughness and geometric anisotropy play important roles. To describe these, molecular dynamics (MD) simulations are used since they consider fluid-solid and fluid-fluid interactions at the molecular level. However, MD simulations are computationally expensive, and are not able to simulate scales larger than a few connected nanopores. We present a method for building and training physics-based deep learning surrogate models to carry out fast and accurate predictions of molecular configurations of gas inside nanopores. Since training deep learning models requires extensive databases that are computationally expensive to create, we employ active learning (AL). AL reduces the overhead of creating comprehensive sets of high-fidelity data by determining where the model uncertainty is greatest, and running simulations on the fly to minimize it. The proposed workflow enables nanoconfinement effects to be rigorously considered at the mesoscale where complex connected sets of nanopores control key applications such as hydrocarbon recovery and CO2 sequestration.

physics.app-ph↗

arXiv2020

Chiral symmetry breaking for deterministic switching of perpendicular magnetization by spin-orbit torque

Symmetry breaking is a characteristic to determine which branch of a bifurcation system follows upon crossing a critical point. Specifically, in spin-orbit torque (SOT) devices, a fundamental question arises: how to break the symmetry of the perpendicular magnetic moment by the in-plane spin polarization? Here, we show that the chiral symmetry breaking by the DMI can induce the deterministic SOT switching of the perpendicular magnetization. By introducing a gradient of saturation magnetization or magnetic anisotropy, non-collinear spin textures are formed by the gradient of effective SOT strength, and thus the chiral symmetry of the SOT-induced spin textures is broken by the DMI, resulting in the deterministic magnetization switching. We introduce a strategy to induce an out-of-plane (z) gradient of magnetic properties, as a practical solution for the wafer-scale manufacture of SOT devices.

cond-mat.mes-hall↗

arXiv2020

Data Age Aware Scheduling for Wireless Powered Mobile-Edge Computing in Industrial Internet of Things

Wireless powered mobile edge computing has been envisioned as a promising paradigm to enhance the computation capability of low-power wireless devices in Industrial Internet of Things. An efficient resource scheduling method is critical yet challenging to design in such a scenario due to stochastic traffic arrival, time-coupling uplink/downlink decision and incomplete system state knowledge. To tackle these challenges, an online optimization algorithm is proposed in this paper to maximize long-term system utility balancing throughput and fairness, subject to data age and stability constraints. A set of virtual queues is designed to transform the scheduling task, which is hard to solve due to time-dependent data age constraints, into a stochastic optimization problem. Leveraging Lyapunov and convex optimization techniques, the proposed approach can achieve asymptotically near-optimal online decisions without any prior statistical knowledge, and maintain the asymptotic optimality in the presence of partial and outdated network state information. Numerical simulations corroborate the theoretical analysis and demonstrate the effectiveness of the proposed approach.

eess.SY↗

arXiv2020

Scaling Limits of Crossing Probabilities in Metric Graph GFF

We consider metric graph Gaussian free field (GFF) defined on polygons of $δ\mathbb{Z}^2$ with alternating boundary data. The crossing probabilities for level-set percolation of metric graph GFF have scaling limits. When the boundary data is well-chosen, the scaling limits of crossing probabilities can be explicitly constructed as "fusion" of multiple SLE$_4$ pure partition functions.

math.PR↗

arXiv2020

Integer Quantization for Deep Learning Inference: Principles and Empirical Evaluation

Quantization techniques can reduce the size of Deep Neural Networks and improve inference latency and throughput by taking advantage of high throughput integer instructions. In this paper we review the mathematical aspects of quantization parameters and evaluate their choices on a wide range of neural network models for different application domains, including vision, speech, and language. We focus on quantization techniques that are amenable to acceleration by processors with high-throughput integer math pipelines. We also present a workflow for 8-bit quantization that is able to maintain accuracy within 1% of the floating-point baseline on all networks studied, including models that are more difficult to quantize, such as MobileNets and BERT-large.

cs.LG↗

arXiv2020

Multivariate Regression of Mixed Responses for Evaluation of Visualization Designs

Information visualization significantly enhances human perception by graphically representing complex data sets. The variety of visualization designs makes it challenging to efficiently evaluate all possible designs catering to users' preferences and characteristics. Most of existing evaluation methods perform user studies to obtain multivariate qualitative responses from users via questionnaires and interviews. However, these methods cannot support online evaluation of designs as they are often time-consuming. A statistical model is desired to predict users' preferences on visualization designs based on non-interference measurements (i.e., wearable sensor signals). In this work, we propose a multivariate regression of mixed responses (MRMR) to facilitate quantitative evaluation of visualization designs. The proposed MRMR method is able to provide accurate model prediction with meaningful variable selection. A simulation study and a user study of evaluating visualization designs with 14 effective participants are conducted to illustrate the merits of the proposed model.

stat.ME↗

arXiv2020

Beyond the Limits of Conventional Stark Deceleration

Stark deceleration enables the production of cold and dense molecular beams with applications in trapping, collisional studies, and precision measurement. Improving the efficiency of Stark deceleration, and hence the achievable molecular densities, is central to unlock the full potential of such studies. One of the chief limitations arises from the transverse focusing properties of Stark decelerators. We introduce a new operation strategy that circumvents this limit without any hardware modifications, and experimentally verify our results for hydroxyl radicals. Notably, improved focusing results in significant gains in molecule yield with increased operating voltage, formerly limited by transverse-longitudinal coupling. At final velocities sufficiently small for trapping, molecule flux improves by a factor of four, and potentially more with increased voltage. The improvement is more significant for less readily polarized species, thereby expanding the class of candidate molecules for Stark deceleration.

physics.chem-ph↗

arXiv2020

Masked Face Recognition Dataset and Application

In order to effectively prevent the spread of COVID-19 virus, almost everyone wears a mask during coronavirus epidemic. This almost makes conventional facial recognition technology ineffective in many cases, such as community access control, face access control, facial attendance, facial security checks at train stations, etc. Therefore, it is very urgent to improve the recognition performance of the existing face recognition technology on the masked faces. Most current advanced face recognition approaches are designed based on deep learning, which depend on a large number of face samples. However, at present, there are no publicly available masked face recognition datasets. To this end, this work proposes three types of masked face datasets, including Masked Face Detection Dataset (MFDD), Real-world Masked Face Recognition Dataset (RMFRD) and Simulated Masked Face Recognition Dataset (SMFRD). Among them, to the best of our knowledge, RMFRD is currently theworld's largest real-world masked face dataset. These datasets are freely available to industry and academia, based on which various applications on masked faces can be developed. The multi-granularity masked face recognition model we developed achieves 95% accuracy, exceeding the results reported by the industry. Our datasets are available at: https://github.com/X-zhangyang/Real-World-Masked-Face-Dataset.

cs.CV↗

arXiv2020

DeepDualMapper: A Gated Fusion Network for Automatic Map Extraction using Aerial Images and Trajectories

Automatic map extraction is of great importance to urban computing and location-based services. Aerial image and GPS trajectory data refer to two different data sources that could be leveraged to generate the map, although they carry different types of information. Most previous works on data fusion between aerial images and data from auxiliary sensors do not fully utilize the information of both modalities and hence suffer from the issue of information loss. We propose a deep convolutional neural network called DeepDualMapper which fuses the aerial image and trajectory data in a more seamless manner to extract the digital map. We design a gated fusion module to explicitly control the information flows from both modalities in a complementary-aware manner. Moreover, we propose a novel densely supervised refinement decoder to generate the prediction in a coarse-to-fine way. Our comprehensive experiments demonstrate that DeepDualMapper can fuse the information of images and trajectories much more effectively than existing approaches, and is able to generate maps with higher accuracy.

cs.CV↗

arXiv2020

Optical Vortex Shaping & Multiple Singularities Manipulation via High-order Cross-phase

Increasing demand for practical applications is forcing deeper research into optical vortices (OVs): from the generation and measurement to shaping and multiple singularities manipulation of OVs. Herein, we propose a new type of phase structure called the high-order cross phase (HOCP) can be employed to modulate OVs to implement both polygonal shaping and singularities manipulation. Theoretically, we investigate the propagation characteristics of OVs with the HOCP. In experiments, we achieve the shaping and singularities manipulation of OVs by utilizing the HOCP. On this basis, we discuss the interference patterns of superposed OVs after the modulation. This work provides an alternative method to achieve both polygonal shaping and multiple singularities manipulation, which will facilitate applications in optical micro-manipulation, optical communication, and high-dimensional quantum entanglement.

physics.optics↗

arXiv2020

Structural-Aware Sentence Similarity with Recursive Optimal Transport

Measuring sentence similarity is a classic topic in natural language processing. Light-weighted similarities are still of particular practical significance even when deep learning models have succeeded in many other tasks. Some light-weighted similarities with more theoretical insights have been demonstrated to be even stronger than supervised deep learning approaches. However, the successful light-weighted models such as Word Mover's Distance [Kusner et al., 2015] or Smooth Inverse Frequency [Arora et al., 2017] failed to detect the difference from the structure of sentences, i.e. order of words. To address this issue, we present Recursive Optimal Transport (ROT) framework to incorporate the structural information with the classic OT. Moreover, we further develop Recursive Optimal Similarity (ROTS) for sentences with the valuable semantic insights from the connections between cosine similarity of weighted average of word vectors and optimal transport. ROTS is structural-aware and with low time complexity compared to optimal transport. Our experiments over 20 sentence textural similarity (STS) datasets show the clear advantage of ROTS over all weakly supervised approaches. Detailed ablation study demonstrate the effectiveness of ROT and the semantic insights.

cs.CL↗

arXiv2020

Crossing estimates from metric graph and discrete GFF

We compare level-set percolation for Gaussian free fields (GFFs) defined on a rectangular subset of $δ\mathbb{Z}^2$ to level-set percolation for GFFs defined on the corresponding metric graph as the mesh size $δ$ goes to 0. In particular, we look at the probability that there is a path that crosses the rectangle in the horizontal direction on which the field is positive. We show this probability is strictly larger in the discrete graph. In the metric graph case, we show that for appropriate boundary conditions the probability that there exists a closed pivotal edge for the horizontal crossing event decays logarithmically in $δ$. In the discrete graph case, we compute the limit of the probability of a horizontal crossing for appropriate boundary conditions.

math.PR↗

arXiv2020

Electric generation from drops impacting onto charged surfaces

The impact of liquid drops onto solid surfaces leads to conversion of kinetic energy of directed drop motion into various forms of energy including surface energy, vibrational energy, heat, and under suitable conditions, electrical energy. The latter has attracted substantial attention in recent years for its potential to directly convert energy from random environmental flows such as rainfall, spray, and wave motion on the sea to electrical energy. Despite the invention of numerous configurations of such energy harvesters, the underlying physical principles and optimum operation conditions have remained elusive. In this letter, we use a combination of high-speed electrical current and video imaging measurements to develop a parameter-free quantitative description of the energy harvesting process for an optimized electrode configuration. A novel electrowetting-assisted charge injection method, EWCI, enables highly stable surface charges and robust energy conversion for several months with record efficiencies exceeding 2.5 percent of the initial kinetic energy.

physics.app-ph↗

arXiv2020

Distributed Brillouin frequency shift extraction via a convolutional neural network

Distributed optical fiber Brillouin sensors detect the temperature and strain along a fiber according to the local Brillouin frequency shift, which is usually calculated by the measured Brillouin spectrum using Lorentzian curve fitting. In addition, cross-correlation, principal component analysis, and machine learning methods have been proposed for the more efficient extraction of Brillouin frequency shifts. However, existing methods only process the Brillouin spectrum individually, ignoring the correlation in the time domain, indicating that there is still room for improvement. Here, we propose and experimentally demonstrate a full convolution neural network to extract the distributed Brillouin frequency shift directly from the measured two-dimensional data. Simulated ideal Brillouin spectrum with various parameters are used to train the network. Both the simulation and experimental results show that the extraction accuracy of the network is better than that of the traditional curve fitting algorithm with a much shorter processing time. This network has good universality and robustness and can effectively improve the performances of existing Brillouin sensors.

eess.SP↗

arXiv2019

Pressure-induced superconductivity in topological type II Dirac semimetal NiTe2

Very recently, NiTe2 has been reported to be a type II Dirac semimetal with Dirac nodes near the Fermi surface. Furthermore, it is unveiled that NiTe2 presents the Hall Effect, which is ascribed to orbital magnetoresistance. The physical properties behavior of NiTe2 under high pressure attracts us. In this paper, we investigate the electrical properties of polycrystalline NiTe2 by application of pressure ranging from 3.4GPa to 54.45Gpa. Superconductivity emerges at critical pressure 12GPa with a transition temperature of 3.7K, and Tc reaches its maximum, 6.4 K, at the pressure of 52.8GPa. Comparing with the superconductivity in MoP, we purposed the possibility of topological superconductivity in NiTe2. Two superconductivity transitions are observed with pressure increasing in single crystal.

cond-mat.supr-con↗

arXiv2019

Picosecond spin-orbit torque switching of ferrimagnets

Spintronics provides an efficient platform for realizing non-volatile memory and logic devices. In these systems, data is stored in the magnetization of magnetic materials, and magnetization is switched in the writing process. In conventional spintronic devices, ferromagnetic materials are used which have a magnetization dynamics timescale of around the nanoseconds, setting a limit for the switching speed. Increasing the magnetization switching speed has been one of the challenges in spintronic research. In this work we take advantage of the ultrafast magnetization dynamics in ferrimagnetic materials instead of ferromagnets, and we use femtosecond laser pulses and a photoconductive Auston switch to create picosecond current pulses for switching the ferrimagnet. By anomalous Hall and magneto-optic Kerr (MOKE) measurement, we demonstrate the robust picosecond SOT driven magnetization switching of ferrimagnetic GdFeCo. The time-resolved MOKE shows more than 50 GHz magnetic resonance frequency of GdFeCo, indicating faster than 20 ps spin dynamics and tens of picosecond SOT switching speed. Our work provides a promising route to realize picosecond operation speed for non-volatile magnetic memory and logic applications.

physics.app-ph↗

arXiv2019

Deep learning Markov and Koopman models with physical constraints

The long-timescale behavior of complex dynamical systems can be described by linear Markov or Koopman models in a suitable latent space. Recent variational approaches allow the latent space representation and the linear dynamical model to be optimized via unsupervised machine learning methods. Incorporation of physical constraints such as time-reversibility or stochasticity into the dynamical model has been established for a linear, but not for arbitrarily nonlinear (deep learning) representations of the latent space. Here we develop theory and methods for deep learning Markov and Koopman models that can bear such physical constraints. We prove that the model is an universal approximator for reversible Markov processes and that it can be optimized with either maximum likelihood or the variational approach of Markov processes (VAMP). We demonstrate that the model performs equally well for equilibrium and systematically better for biased data compared to existing approaches, thus providing a tool to study the long-timescale processes of dynamical systems.

physics.comp-ph↗