arXiv subjects
Publications and source records attributed to Chao Li.
In this paper, we consider the Dirichlet problem of a complex Monge-Ampère equation on a ball in $\mathbb C^n$. With $\mathcal C^{1,α}$ (resp. $\mathcal C^{0,α}$) data, we prove an interior $\mathcal C^{1,α}$ (resp. $\mathcal C^{0,α}$) estimate for the solution. These estimates are generalized versions of the Bedford-Taylor interior $\mathcal C^{1,1}$ estimate.
We study positive scalar curvature on the regular part of Riemannian manifolds with singular, uniformly Euclidean ($L^\infty$) metrics that consolidate Gromov's scalar curvature polyhedral comparison theory and edge metrics that appear in the study of Einstein manifolds. We show that, in all dimensions, edge singularities with cone angles $\leq 2π$ along codimension-2 submanifolds do not affect the Yamabe type. In three dimensions, we prove the same for more general singular sets, which are allowed to stratify along 1-skeletons, exhibiting edge singularities (angles $\leq 2π$) and arbitrary $L^\infty$ isolated point singularities. We derive, as an application of our techniques, Positive Mass Theorems for asymptotically flat manifolds with analogous singularities.
Ubiquitous deployment of low-cost smart devices and widespread use of high-speed wireless networks have led to the rapid development of the Internet of Things (IoT). IoT embraces countless physical objects that have not been involved in the traditional Internet and enables their interaction and cooperation to provide a wide range of IoT applications. Many services in the IoT may require a comprehensive understanding and analysis of data collected through a large number of physical devices that challenges both personal information privacy and the development of IoT. Information privacy in IoT is a broad and complex concept as its understanding and perception differ among individuals and its enforcement requires efforts from both legislation as well as technologies. In this paper, we review the state-of-the-art principles of privacy laws, the architectures for IoT and the representative privacy enhancing technologies (PETs). We analyze how legal principles can be supported through a careful implementation of privacy enhancing technologies (PETs) at various layers of a layered IoT architecture model to meet the privacy requirements of the individuals interacting with IoT systems. We demonstrate how privacy legislation maps to privacy principles which in turn drives the design of necessary privacy enhancing technologies to be employed in the IoT architecture stack.
Understanding the impact of lattice imperfections on nanoscale thermal transport is crucial for diverse applications ranging from thermal management to energy conversion. Grain boundaries (GBs) are ubiquitous defects in polycrystalline materials, which scatter phonons and reduce thermal conductivity. Historically, their impact on heat conduction has been studied indirectly through spatially-averaged measurements, that provide little information about phonon transport near a single GB. Here, using spatially-resolved time-domain thermoreflectance (TDTR) measurements in combination with electron backscatter diffraction (EBSD), we make localized measurements of thermal conductivity within few μm of individual GBs in boron-doped polycrystalline diamond. We observe strongly suppressed thermal transport near GBs, a reduction in conductivity from ~1000 W/m-K at the center of large grains to ~400 W/m-K in the immediate vicinity of GBs. Furthermore, we show that this reduction in conductivity is measured up to ~10 μm away from a GB. A theoretical model is proposed that captures the local reduction in phonon mean-free-paths due to strongly diffuse phonon scattering at the disordered grain boundaries. Our results provide a new framework for understanding phonon-defect interactions in nanomaterials, with implications for the use of high thermal conductivity polycrystalline materials as heat sinks in electronics thermal management.
Compared with visible object tracking, thermal infrared (TIR) object tracking can track an arbitrary target in total darkness since it cannot be influenced by illumination variations. However, there are many unwanted attributes that constrain the potentials of TIR tracking, such as the absence of visual color patterns and low resolutions. Recently, structured output support vector machine (SOSVM) and discriminative correlation filter (DCF) have been successfully applied to visible object tracking, respectively. Motivated by these, in this paper, we propose a large margin structured convolution operator (LMSCO) to achieve efficient TIR object tracking. To improve the tracking performance, we employ the spatial regularization and implicit interpolation to obtain continuous deep feature maps, including deep appearance features and deep motion features, of the TIR targets. Finally, a collaborative optimization strategy is exploited to significantly update the operators. Our approach not only inherits the advantage of the strong discriminative capability of SOSVM but also achieves accurate and robust tracking with higher-dimensional features and more dense samples. To the best of our knowledge, we are the first to incorporate the advantages of DCF and SOSVM for TIR object tracking. Comprehensive evaluations on two thermal infrared tracking benchmarks, i.e. VOT-TIR2015 and VOT-TIR2016, clearly demonstrate that our LMSCO tracker achieves impressive results and outperforms most state-of-the-art trackers in terms of accuracy and robustness with sufficient frame rate.
Discriminative Correlation Filters based tracking algorithms exploiting conventional handcrafted features have achieved impressive results both in terms of accuracy and robustness. Template handcrafted features have shown excellent performance, but they perform poorly when the appearance of target changes rapidly such as fast motions and fast deformations. In contrast, statistical handcrafted features are insensitive to fast states changes, but they yield inferior performance in the scenarios of illumination variations and background clutters. In this work, to achieve an efficient tracking performance, we propose a novel visual tracking algorithm, named MFCMT, based on a complementary ensemble model with multiple features, including Histogram of Oriented Gradients (HOGs), Color Names (CNs) and Color Histograms (CHs). Additionally, to improve tracking results and prevent targets drift, we introduce an effective fusion method by exploiting relative entropy to coalesce all basic response maps and get an optimal response. Furthermore, we suggest a simple but efficient update strategy to boost tracking performance. Comprehensive evaluations are conducted on two tracking benchmarks demonstrate and the experimental results demonstrate that our method is competitive with numerous state-of-the-art trackers. Our tracker achieves impressive performance with faster speed on these benchmarks.
This paper proposes a real-time dynamic scene reconstruction method capable of reproducing the motion, geometry, and segmentation simultaneously given live depth stream from a single RGB-D camera. Our approach fuses geometry frame by frame and uses a segmentation-enhanced node graph structure to drive the deformation of geometry in registration step. A two-level node motion optimization is proposed. The optimization space of node motions and the range of physically-plausible deformations are largely reduced by taking advantage of the articulated motion prior, which is solved by an efficient node graph segmentation method. Compared to previous fusion-based dynamic scene reconstruction methods, our experiments show robust and improved reconstruction results for tangential and occluded motions.
In particle accelerators, particle losses to surrounding components must be minimized. These particles usually have been driven into an undesired "halo" outside the desired distribution before being lost. The processes involved and physical identification of such halo is an intense research topic especially when the nonlinear space charge effect is taken into account. In order to present a method, firmly based on the physics of the transient beam state, for determining when a particle may be defined as being in a halo, the particular example of an initially rms matched beam injected into a resonance of an alternating-gradient (AG) focusing lattice is studied, in which the resonance between the collective mode and single particle motion pushes some particles into a halo defined by extending the halo description to the action-angle frame. In this new frame of description, the halo particles can be physically defined as those with larger actions affected by the outer quasi hyperbolic tori, which contains the halo particles observed in the traditional approaches on projected 2D phase space, as well as some particles still hidden in the core.
As an extension of previous theoretical study on the coherent structure resonance due to space charge effects [Chao Li and R. A. Jameson, Phys. Rev. Accel. Beams 21, 024204, 2018], this paper aims to demonstrate how the beam, as a whole, is spontaneously affected when the predicted mixed coherent 2nd/4th order structure resonance stop band around $90^{\circ}$ phase advance is crossed. The beam characteristics during the structure resonance crossing, such as the rms emittance growth and the appearance of 2-fold/4-fold structure in phase space, are well explained by the mixture characteristic of different orders of structure resonance. The related `attracting' and `repulsive' effects in the structure resonance stop band from below and above crossing are considered as a natural beam reaction to the coherent structure resonance that the beam spontaneously moves to a structure resonance free region then the space charge takes a weaker importance. In the PIC simulation, it is found that the emittance growth is positively related to the time that the beam spends inside the structure resonance stop band. As a potential candidate mechanism, the incoherent particle-core resonance also has been checked. It is found that this incoherent resonance has basic difficulties in explaining the results obtained from the self-consistent PIC simulation. The incoherent particle-core resonance might lead to phase space distortion, emittance growth, or beam halo formation only on a long-time scale. This clarity of the discrepancies between coherent and incoherent resonance mechanisms will lead to a better understanding of the numerical study and experimental researches obtained recently. In addition, similar understanding can be extended to the study of higher order structure resonance.
The Tianlai Pathfinder is designed to demonstrate the feasibility of using a wide field of view radio interferometers to map the density of neutral hydrogen in the Universe after the Epoch of Reionizaton. This approach, called 21~cm intensity-mapping, promises an inexpensive means for surveying the large-scale structure of the cosmos. The Tianlai Pathfinder presently consists of an array of three, 15~m $\times$ 40~m cylinder telescopes and an array of sixteen, 6~m diameter dish antennas located in a radio-quiet part of western China. The two types of arrays were chosen to determine the advantages and disadvantages of each approach. The primary goal of the Pathfinder is to make 3D maps by surveying neutral hydrogen over large areas of the sky %$20,000 {\rm deg}^2$ in two different redshift ranges: first at $1.03 > z > 0.78$ ($700 - 800$~MHz) and later at $0.21 > z > 0.12$ ($1170 - 1270$~MHz). The most significant challenge to $21$~cm intensity-mapping is the removal of strong foreground radiation that dwarfs the cosmological signal. It requires exquisite knowledge of the instrumental response, i.e. calibration. In this paper, we provide an overview of the status of the Pathfinder and discuss the details of some of the analysis that we have carried out to measure the beam function of both arrays. We compare electromagnetic simulations of the arrays to measurements, discuss measurements of the gain and phase stability of the instrument, and provide a brief overview of the data processing pipeline.
The Compressed Baryonic Matter (CBM) experiment will investigate the quantum chromodynamics (QCD) phase diagram at high net baryon densities and moderate temperatures. CBM Time of Flight (TOF) system is composed of super modules containing high performance Multi-gap Resistive Plate Chambers (MRPCs). During the mass production, each super module assembled with MRPCs needs quality evaluation, which includes time measurement and data readout. Read out electronics encounter the challenge of reading data from a super module at a speed of about 6 Gbps. In this paper, a read out method based on Peripheral Component Interconnect Express (PCIe) over optical fiber is proposed for CBM-TOF super module quality evaluation. The digitized data from super module will be concentrated at the front-end electronics, and then be transmitted to a PCIe switch module (PSM) over optical fiber using PCIe protocol. The PSM is directly plugged into the motherboard via gold fingers at the backend data acquisition server. With this readout method, a high-speed transmission rate can be reached. Furthermore, a PSM can receives data from several super modules simultaneously, which is important to improve the evaluation efficiency. This readout method simplifies the architecture of readout electronics and supports long distance transmission between frontend and backend.
The time-of-flight (TOF) system in the Compressed Baryonic Matter (CBM) experiment is composed of super modules based on multi-gap Resistive Plate Chambers (MRPC) for high-denseness, high-resolution time measurement. In order to evaluate the quality of detectors during the mass production, a distributed data readout system is developed to meet with the high data rate of 6.4Gbps, and a related data acquisition (DAQ) software is implemented under Linux operating system to test and verify the feasibility of the distributed data readout method. In this paper, the DAQ software is focused on data collection, event building, status monitoring and system controlling. Laboratory tests confirmed the function of the DAQ software and show that the overall data transfer rate of a single data transmission path can reach up to about 550Mbps which already meet the demand.
It is well known that the Euler vortex patch in $\mathbb{R}^{2}$ will remain regular if it is regular enough initially. In bounded domains, the regularity theory for patch solutions is less complete. In this paper, we study Euler vortex patches in a general smooth bounded domain. We prove global in time regularity by providing an upper bound on the growth of curvature of the patch boundary. For a special symmetric scenario, we construct an example of double exponential curvature growth, showing that our upper bound is qualitatively sharp.
In low-rank tensor completion tasks, due to the underlying multiple large-scale singular value decomposition (SVD) operations and rank selection problem of the traditional methods, they suffer from high computational cost and high sensitivity of model complexity. In this paper, taking advantages of high compressibility of the recently proposed tensor ring (TR) decomposition, we propose a new model for tensor completion problem. This is achieved through introducing convex surrogates of tensor low-rank assumption on latent tensor ring factors, which makes it possible for the Schatten norm regularization based models to be solved at much smaller scale. We propose two algorithms which apply different structured Schatten norms on tensor ring factors respectively. By the alternating direction method of multipliers (ADMM) scheme, the tensor ring factors and the predicted tensor can be optimized simultaneously. The experiments on synthetic data and real-world data show the high performance and efficiency of the proposed approach.
Skeleton-based human action recognition has recently drawn increasing attentions with the availability of large-scale skeleton datasets. The most crucial factors for this task lie in two aspects: the intra-frame representation for joint co-occurrences and the inter-frame representation for skeletons' temporal evolutions. In this paper we propose an end-to-end convolutional co-occurrence feature learning framework. The co-occurrence features are learned with a hierarchical methodology, in which different levels of contextual information are aggregated gradually. Firstly point-level information of each joint is encoded independently. Then they are assembled into semantic representation in both spatial and temporal domains. Specifically, we introduce a global spatial aggregation scheme, which is able to learn superior joint co-occurrence features over local aggregation. Besides, raw skeleton coordinates as well as their temporal difference are integrated with a two-stream paradigm. Experiments show that our approach consistently outperforms other state-of-the-arts on action recognition and detection benchmarks like NTU RGB+D, SBU Kinect Interaction and PKU-MMD.
High-dimensional data in many areas such as computer vision and machine learning tasks brings in computational and analytical difficulty. Feature selection which selects a subset from observed features is a widely used approach for improving performance and effectiveness of machine learning models with high-dimensional data. In this paper, we propose a novel AutoEncoder Feature Selector (AEFS) for unsupervised feature selection which combines autoencoder regression and group lasso tasks. Compared to traditional feature selection methods, AEFS can select the most important features by excavating both linear and nonlinear information among features, which is more flexible than the conventional self-representation method for unsupervised feature selection with only linear assumptions. Experimental results on benchmark dataset show that the proposed method is superior to the state-of-the-art method.
In this work, we consider the task of classifying binary positive-unlabeled (PU) data. The existing discriminative learning based PU models attempt to seek an optimal reweighting strategy for U data, so that a decent decision boundary can be found. However, given limited P data, the conventional PU models tend to suffer from overfitting when adapted to very flexible deep neural networks. In contrast, we are the first to innovate a totally new paradigm to attack the binary PU task, from perspective of generative learning by leveraging the powerful generative adversarial networks (GAN). Our generative positive-unlabeled (GenPU) framework incorporates an array of discriminators and generators that are endowed with different roles in simultaneously producing positive and negative realistic samples. We provide theoretical analysis to justify that, at equilibrium, GenPU is capable of recovering both positive and negative data distributions. Moreover, we show GenPU is generalizable and closely related to the semi-supervised classification. Given rather limited P data, experiments on both synthetic and real-world dataset demonstrate the effectiveness of our proposed framework. With infinite realistic and diverse sample streams generated from GenPU, a very flexible classifier can then be trained using deep neural networks.
Thanks to the success of deep learning, cross-modal retrieval has made significant progress recently. However, there still remains a crucial bottleneck: how to bridge the modality gap to further enhance the retrieval accuracy. In this paper, we propose a self-supervised adversarial hashing (\textbf{SSAH}) approach, which lies among the early attempts to incorporate adversarial learning into cross-modal hashing in a self-supervised fashion. The primary contribution of this work is that two adversarial networks are leveraged to maximize the semantic correlation and consistency of the representations between different modalities. In addition, we harness a self-supervised semantic network to discover high-level semantic information in the form of multi-label annotations. Such information guides the feature learning process and preserves the modality relationships in both the common semantic space and the Hamming space. Extensive experiments carried out on three benchmark datasets validate that the proposed SSAH surpasses the state-of-the-art methods.