arXiv subjects

Juan Du

Publications and source records attributed to Juan Du.

50 records · Page 3Linked to original sources

arXiv2021

Reinforcement Learning for Process Control with Application in Semiconductor Manufacturing

Process control is widely discussed in the manufacturing process, especially for semiconductor manufacturing. Due to unavoidable disturbances in manufacturing, different process controllers are proposed to realize variation reduction. Since reinforcement learning (RL) has shown great advantages in learning actions from interactions with a dynamic system, we introduce RL methods for process control and propose a new controller called RL-based controller. Considering the fact that most existing process controllers mainly rely on a linear model assumption for the process input-output relationship, we first propose theoretical properties of RL-based controllers according to the linear model assumption. Then the performance of RL-based controllers and traditional process controllers (e.g., exponentially weighted moving average (EWMA), general harmonic rule (GHR) controllers) are compared for linear processes. Furthermore, we find that the RL-based controllers have potential advantages to deal with other complicated nonlinear processes that are with and without assumed explicit model formulations. The intensive numerical studies validate the advantages of the proposed RL-based controllers.

eess.SY↗

arXiv2021

Acoustic Möbius insulators from projective symmetry

Symmetry plays a critical role in classifying phases of matter. This is exemplified by how crystalline symmetries enrich the topological classification of materials and enable unconventional phenomena in topologically nontrivial ones. After an extensive study over the past decade, the list of topological crystalline insulators and semimetals seems to be exhaustive and concluded. However, in the presence of gauge symmetry, common but not limited to artificial crystals, the algebraic structure of crystalline symmetries needs to be projectively represented, giving rise to unprecedented topological physics. Here we demonstrate this novel idea by exploiting a projective translation symmetry and constructing a variety of Möbius-twisted topological phases. Experimentally, we realize two Möbius insulators in acoustic crystals for the first time: a two-dimensional one of first-order band topology and a three-dimensional one of higher-order band topology. We observe unambiguously the peculiar Möbius edge and hinge states via real-space visualization of their localiztions, momentum-space spectroscopy of their 4π periodicity, and phase-space winding of their projective translation eigenvalues. Not only does our work open a new avenue for artificial systems under the interplay between gauge and crystalline symmetries, but it also initializes a new framework for topological physics from projective symmetry.

cond-mat.mes-hall↗

arXiv2021

Disorder-assisted Robustness of Ultrafast Cooling in High Doped CVD-Graphene

Dirac Fermion, which is the low energy collective excitation near the Dirac cone in monolayer graphene, have gained great attention by low energy Terahertz probe. In the case of undoped graphene, it has been generally understood that the ultrafast terahertz thermal relaxation is mostly driven by the electron-phonon coupling (EOP), which can be prolonged to tens and hundreds of picoseconds. However, for the high doped graphene, which manifests the negative photoinduced terahertz conductivity, there is still no consensus on the dominant aspects of the cooling process on a time scale of a few picoseconds. Here, the competition between the disorders assisted defect scattering and the electron-phonon coupling process in the cooling process of the graphene terahertz dynamics is systematically studied and disentangled. We verify experimentally that the ultrafast disorder assisted lattice-phonon interaction, rather than the electron-phonon coupling process, would play the key role in the ultrafast thermal relaxation of the terahertz dynamics. Furthermore, the cooling process features robustness which is independent on the pump wavelength and external temperature. Our finding is expected to propose a considerable possible cooling channel in CVD-graphene and to increase the hot electron extracting efficiency for the design of graphene-based photoconversion devices.

cond-mat.mes-hall↗

arXiv2021

SOE-Net: A Self-Attention and Orientation Encoding Network for Point Cloud based Place Recognition

We tackle the problem of place recognition from point cloud data and introduce a self-attention and orientation encoding network (SOE-Net) that fully explores the relationship between points and incorporates long-range context into point-wise local descriptors. Local information of each point from eight orientations is captured in a PointOE module, whereas long-range feature dependencies among local descriptors are captured with a self-attention unit. Moreover, we propose a novel loss function called Hard Positive Hard Negative quadruplet loss (HPHN quadruplet), that achieves better performance than the commonly used metric learning loss. Experiments on various benchmark datasets demonstrate superior performance of the proposed network over the current state-of-the-art approaches. Our code is released publicly at https://github.com/Yan-Xia/SOE-Net.

cs.CV↗

arXiv2021

Topological dislocation modes in three-dimensional acoustic topological insulators

Dislocations are ubiquitous in three-dimensional solid-state materials. The interplay of such real space topology with the emergent band topology defined in reciprocal space gives rise to gapless helical modes bound to the line defects. This is known as bulk-dislocation correspondence, in contrast to the conventional bulk-boundary correspondence featuring topological states at boundaries. However, to date rare compelling experimental evidences are presented for this intriguing topological observable, owing to the presence of various challenges in solid-state systems. Here, using a three-dimensional acoustic topological insulator with precisely controllable dislocations, we report an unambiguous experimental evidence for the long-desired bulk-dislocation correspondence, through directly measuring the gapless dispersion of the one-dimensional topological dislocation modes. Remarkably, as revealed in our further experiments, the pseudospin-locked dislocation modes can be unidirectionally guided in an arbitrarily-shaped dislocation path. The peculiar topological dislocation transport, expected in a variety of classical wave systems, can provide unprecedented controllability over wave propagations.

cond-mat.mes-hall↗

arXiv2020

Quantitative understanding of negative thermal expansion in scandium trifluoride from neutron total scattering measurements

Negative thermal expansion (NTE) - the phenomenon where some materials shrink rather than expand when heated - is both intriguing and useful, but remains poorly understood. Current understanding hinges on the role of specific vibrational modes, but in fact thermal expansion is a weighted sum of contributions from every possible mode. Here we overcome this difficulty by deriving a real-space model of atomic motion in the prototypical NTE material scandium trifluoride, ScF3, from total neutron scattering data. We show that NTE in this material depends not only on rigid unit modes - the vibrations in which the scandium coordination octahedra remain undistorted - but also on modes that distort these octahedra. Furthermore, in contrast with previous predictions, we show that the quasiharmonic approximation coupled with renormalisation through anharmonic interactions describes this behaviour well. Our results point the way towards a new understanding of how NTE is manifested in real materials.

cond-mat.mtrl-sci↗

arXiv2020

DH3D: Deep Hierarchical 3D Descriptors for Robust Large-Scale 6DoF Relocalization

For relocalization in large-scale point clouds, we propose the first approach that unifies global place recognition and local 6DoF pose refinement. To this end, we design a Siamese network that jointly learns 3D local feature detection and description directly from raw 3D points. It integrates FlexConv and Squeeze-and-Excitation (SE) to assure that the learned local descriptor captures multi-level geometric information and channel-wise relations. For detecting 3D keypoints we predict the discriminativeness of the local descriptors in an unsupervised manner. We generate the global descriptor by directly aggregating the learned local descriptors with an effective attention mechanism. In this way, local and global 3D descriptors are inferred in one single forward pass. Experiments on various benchmarks demonstrate that our method achieves competitive results for both global point cloud retrieval and local point cloud registration in comparison to state-of-the-art approaches. To validate the generalizability and robustness of our 3D keypoints, we demonstrate that our method also performs favorably without fine-tuning on the registration of point clouds that were generated by a visual SLAM system. Code and related materials are available at https://vision.in.tum.de/research/vslam/dh3d.

cs.CV↗

arXiv2019

Optimal Shape Control via $L_\infty$ Loss for Composite Fuselage Assembly

Shape control is critical to ensure the quality of composite fuselage assembly. In current practice, the structures are adjusted to the design shape in terms of the $\ell_2$ loss for further assembly without considering the existing dimensional gap between two structures. Such practice has two limitations: (1) the design shape may not be the optimal shape in terms of a pair of incoming fuselages with different incoming dimensions; (2) the maximum gap is the key concern during the fuselage assembly process. This paper proposes an optimal shape control methodology via the $\ell_\infty$ loss for composite fuselage assembly process by considering the existing dimensional gap between the incoming pair of fuselages. Besides, due to the limitation on the number of available actuators in practice, we face an important problem of finding the best locations for the actuators among many potential locations, which makes the problem a sparse estimation problem. We are the first to solve the optimal shape control in fuselage assembly process using the $\ell_\infty$ model under the framework of sparse estimation, where we use the $\ell_1$ penalty to control the sparsity of the resulting estimator. From statistical point of view, this can be formulated as the $\ell_\infty$ loss based linear regression, and under some standard assumptions, such as the restricted eigenvalue (RE) conditions, and the light tailed noise, the non-asymptotic estimation error of the $\ell_1$ regularized $\ell_\infty$ linear model is derived to be the order of $O(σ\sqrt{\frac{S\log p}{n}})$, which meets the upper-bound in the existing literature. Compared to the current practice, the case study shows that our proposed method significantly reduces the maximum gap between two fuselages after shape adjustments.

stat.AP↗

arXiv2019

Landscape of Big Medical Data: A Pragmatic Survey on Prioritized Tasks

Big medical data poses great challenges to life scientists, clinicians, computer scientists, and engineers. In this paper, a group of life scientists, clinicians, computer scientists and engineers sit together to discuss several fundamental issues. First, what are the unique characteristics of big medical data different from those of the other domains? Second, what are the prioritized tasks in clinician research and practices utilizing big medical data? And do we have enough publicly available data sets for performing those tasks? Third, do the state-of-the-practice and state-of-the-art algorithms perform good jobs? Fourth, are there any benchmarks for measuring algorithms and systems for big medical data? Fifth, what are the performance gaps of state-of-the-practice and state-of-the-art systems handling big medical data currently or in future? Finally but not least, are we, life scientists, clinicians, computer scientists and engineers, ready for working together? We believe answering the above issues will help define and shape the landscape of big medical data.

cs.CY↗

arXiv2018

Object Tracking in Hyperspectral Videos with Convolutional Features and Kernelized Correlation Filter

Target tracking in hyperspectral videos is a new research topic. In this paper, a novel method based on convolutional network and Kernelized Correlation Filter (KCF) framework is presented for tracking objects of interest in hyperspectral videos. We extract a set of normalized three-dimensional cubes from the target region as fixed convolution filters which contain spectral information surrounding a target. The feature maps generated by convolutional operations are combined to form a three-dimensional representation of an object, thereby providing effective encoding of local spectral-spatial information. We show that a simple two-layer convolutional networks is sufficient to learn robust representations without the need of offline training with a large dataset. In the tracking step, KCF is adopted to distinguish targets from neighboring environment. Experimental results demonstrate that the proposed method performs well on sample hyperspectral videos, and outperforms several state-of-the-art methods tested on grayscale and color videos in the same scene.

cs.CV↗

arXiv2015

Cyclotron Dynamics of a Kondo Singlet in a Spin-Orbit-Coupled Alkaline-Earth Atomic Gas

We propose a scheme to investigate the interplay between Kondo-exchange interaction and quantum spin Hall effect with ultracold fermionic alkaline-earth atoms trapped in two-dimensional optical lattices using ultracold collision and laser-assisted tunneling. In the strong Kondo-coupling regime, though the loop trajectory of the mobile atom disappears, collective dynamics of an atom pair in two clock states can exhibit an unexpected spin-dependent cyclotron orbit in a plaquette, realizing the quantum spin Hall effect of the Kondo singlet. We demonstrate that the collective cyclotron dynamics of the spin-zero Kondo singlet is governed by an effective Harper-Hofstadter model in addition to second-order diagonal tunneling.

cond-mat.quant-gas↗

arXiv2014

Interacting heavy fermions in a disordered optical lattice

We have theoretically studied the effect of disorder on ultracold alkaline-earth atoms governed by the Kondo lattice model in an optical lattice via simplified double-well model and hybridization mean-field theory. Disorder-induced narrowing and even complete closure of hybridization gap have been predicted and the compressibility of the system has also been investigated for metallic and Kondo insulator phases in the presence of the disordered potential. To make connection to the experimental situation, we have numerically solved the disordered Kondo lattice model with an external harmonic trap and shown both the melting of Kondo insulator plateau and an compressibility anomaly at low-density.

cond-mat.quant-gas↗

arXiv2009

Anisotropic Magneto-conductance of InAs Nanowire: Angle Dependent Suppression of 1D Weak Localization

The magneto-conductance of an InAs nanowire is investigated with respect to the relative orientation between external magnetic field and the nanowire axis. It is found that both the perpendicular and the parallel magnetic fields induce a positive magneto-conductance. Yet the parallel magnetic field induced longitudinal magneto-conductance has a smaller magnitude. This anisotropic magneto-transport phenomenon is studied as a function of temperature, magnetic field strength and at an arbitrary angle between the magnetic field and the nanowire. We show that the observed effect is in quantitative agreement with the suppression of one-dimensional (1D) weak localization.

cond-mat.mes-hall↗

arXiv2009

Fixed-domain asymptotic properties of tapered maximum likelihood estimators

When the spatial sample size is extremely large, which occurs in many environmental and ecological studies, operations on the large covariance matrix are a numerical challenge. Covariance tapering is a technique to alleviate the numerical challenges. Under the assumption that data are collected along a line in a bounded region, we investigate how the tapering affects the asymptotic efficiency of the maximum likelihood estimator (MLE) for the microergodic parameter in the Matérn covariance function by establishing the fixed-domain asymptotic distribution of the exact MLE and that of the tapered MLE. Our results imply that, under some conditions on the taper, the tapered MLE is asymptotically as efficient as the true MLE for the microergodic parameter in the Matérn model.

math.ST↗