arXiv subjects

Xinyang Li

Publications and source records attributed to Xinyang Li.

85 records · Page 5Linked to original sources

arXiv2023

Monitoring the evolution of dimensional accuracy and product properties in property-controlled forming processes

As recent trends in manufacturing engineering disciplines show a clear development in the sustainable as well as economically efficient design of forming processes, monitoring techniques have been gaining in relevance. In terms of monitoring of product properties, most processes are currently open-loop controlled, entailing that the microstructure evolution, which determines the final product properties, is not considered. However, a closed-loop control that can adjust and manipulate the process actuators according to the required product properties of the component will lead to a considerable increase in efficiency of the processes regarding resources and will decrease postproduction of the component. For most forming processes, one set of component dimensions will result in a certain set of product properties. However, to successfully establish closed-loop property controls for the processes, a systematic understanding of the reciprocity of the dimensions after forming and final product properties must be established. This work investigates the evolution of dimensional accuracy as well as product properties for a series of forming processes that utilize different degrees of freedom for process control.

eess.SY↗

arXiv2023

Sensing-Assisted Receivers for Resilient-By-Design 6G MU-MIMO Uplink

We address the resilience of future 6G MIMO communications by considering an uplink scenario where multiple legitimate transmitters try to communicate with a base station in the presence of an adversarial jammer. The jammer possesses full knowledge about the system and the physical parameters of the legitimate link, while the base station only knows the UL-channels and the angle-of-arrival (AoA) of the jamming signals. Furthermore, the legitimate transmitters are oblivious to the fact that jamming takes place, thus the burden of guaranteeing resilience falls on the receiver. For this case we derive one optimal jamming strategy that aims to minimize the rate of the strongest user and multiple receive strategies, one based on a lower bound on the achievable signal-to-interference-to-noise-ratio (SINR), one based on a zero-forcing (ZF) design, and one based on a minimum SINR constraint. Numerical studies show that the proposed anti-jamming approaches ensure that the sum rate of the system is much higher than without protection, even when the jammer has considerably more transmit power and even if the jamming signals come from the same direction as those of the legitimate users.

eess.SP↗

arXiv2023

Optimal and Robust Waveform Design for MIMO-OFDM Channel Sensing: A Cramér-Rao Bound Perspective

Wireless channel sensing is one of the key enablers for integrated sensing and communication (ISAC) which helps communication networks understand the surrounding environment. In this work, we consider MIMO-OFDM systems and aim to design optimal and robust waveforms for accurate channel parameter estimation given allocated OFDM resources. The Fisher information matrix (FIM) is derived first, and the waveform design problem is formulated by maximizing the log determinant of the FIM. We then consider the uncertainty in the parameters and state the stochastic optimization problem for a robust design. We propose the Riemannian Exact Penalty Method via Smoothing (REPMS) and its stochastic version SREPMS to solve the constrained non-convex problems. In simulations, we show that the REPMS yields comparable results to the semidefinite relaxation (SDR) but with a much shorter running time. Finally, the designed robust waveforms using SREMPS are investigated, and are shown to have a good performance under channel perturbations.

cs.IT↗

arXiv2022

Machine learning glass caging order parameters with an artificial nested neural network

Around a glass transition, the dynamics of a supercooled liquid dramatically slow down, exhibited by caging of particles, while the structural changes remain subtle. In alternative to recent machine learning studies searching for structural predictors of glassy dynamics, here we propose to learn directly particle caging features defined purely according to dynamics. We focus on three transitions in a simulated hard sphere glass model, the melting of ultra-stable glasses, the Gardner transition and the liquid to ordinary glass transition. Implementing the machine learning algorithm based on a two-level nested neural network, we attain not only proper caging order parameters for all three transitions, but also a phase classification for input samples. A finite-size scaling analysis of the phase classification results identifies the order of melting (first) and Gardner (second) transitions. A false positive is avoided, as the liquid to glass transition is indicated as a crossover, rather than a phase transition with a well-defined transition point. This study paves the way to a generic approach for learning dynamical features in glassy systems, with a minimum requirement of system-specific knowledge.

cond-mat.dis-nn↗

arXiv2021

Theory of Vibrational Polariton Chemistry in the Collective Coupling Regime

We theoretically demonstrate that chemical reaction rate constant can be significantly suppressed by coupling molecular vibrations with an optical cavity, exhibiting both the collective coupling effect and the cavity-frequency modification of the rate constant. When a reaction coordinate is strongly coupled to the solvent molecules, the reaction rate constant is reduced due to the dynamical caging effect. We demonstrate that collectively coupling the solvent to the cavity can further enhance this dynamical caging effect, leading to additional suppression of the chemical kinetics. This effect is further amplified when cavity loss is considered.

physics.chem-ph↗

arXiv2021

Deep learning piston aberration control of fiber laser phased array by spiral phase modulation

The stochastic parallel gradient descent (SPGD) algorithm is usually employed as the control strategy for phase-locking in fiber laser phased array systems. However, the convergence speed of the SPGD algorithm will slow down as the number of array elements increases. To improve the control bandwidth, the convolutional neural network is introduced to quickly calculate the initial piston aberration in a single step. In addition, the irrationality of the commonly used Mean Square Error (MSE) evaluation function in existing convolutional neural networks is analyzed. A new evaluation function NPCD (Normalized Phase Cosine Distance) is proposed to improve the accuracy of the neural networks. The results show that the piston aberration residual is 0.005 and the power in the bucket (PIB) is 0.993 after accurate preliminary compensation, which means that the system directly enters the co-phase state. We also demonstrate the robustness and scalability by adding additional disturbance and expanding the scale of the array.

physics.optics↗

arXiv2021

Prostate cancer histopathology with label-free multispectral deep UV microscopy quantifies phenotypes of tumor grade and aggressiveness

Identifying prostate cancer patients that are harboring aggressive forms of prostate cancer remains a significant clinical challenge. To shed light on this problem, we develop an approach based on multispectral deep-ultraviolet (UV) microscopy that provides novel quantitative insight into the aggressiveness and grade of this disease. First, we find that UV spectral signatures from endogenous molecules give rise to a phenotypical continuum that differentiates critical structures of thin tissue sections with subcellular spatial resolution, including nuclei, cytoplasm, stroma, basal cells, nerves, and inflammation. Further, we show that this phenotypical continuum can be applied as a surrogate biomarker of prostate cancer malignancy, where patients with the most aggressive tumors show a ubiquitous glandular phenotypical shift. Lastly, we adapt a two-part Cycle-consistent Generative Adversarial Network to translate the label-free deep-UV images into virtual hematoxylin and eosin (H&E) stained images. Agreement between the virtual H&E images and the gold standard H&E-stained tissue sections is evaluated by a panel of pathologists who find that the two modalities are in excellent agreement. This work has significant implications towards improving our ability to objectively quantify prostate cancer grade and aggressiveness, thus improving the management and clinical outcomes of prostate cancer patients. This same approach can also be applied broadly in other tumor types to achieve low-cost, stain-free, quantitative histopathological analysis.

physics.med-ph↗

arXiv2021

Image-to-image Translation via Hierarchical Style Disentanglement

Recently, image-to-image translation has made significant progress in achieving both multi-label (\ie, translation conditioned on different labels) and multi-style (\ie, generation with diverse styles) tasks. However, due to the unexplored independence and exclusiveness in the labels, existing endeavors are defeated by involving uncontrolled manipulations to the translation results. In this paper, we propose Hierarchical Style Disentanglement (HiSD) to address this issue. Specifically, we organize the labels into a hierarchical tree structure, in which independent tags, exclusive attributes, and disentangled styles are allocated from top to bottom. Correspondingly, a new translation process is designed to adapt the above structure, in which the styles are identified for controllable translations. Both qualitative and quantitative results on the CelebA-HQ dataset verify the ability of the proposed HiSD. We hope our method will serve as a solid baseline and provide fresh insights with the hierarchically organized annotations for future research in image-to-image translation. The code has been released at https://github.com/imlixinyang/HiSD.

cs.CV↗

arXiv2020

Keypoint Autoencoders: Learning Interest Points of Semantics

Understanding point clouds is of great importance. Many previous methods focus on detecting salient keypoints to identity structures of point clouds. However, existing methods neglect the semantics of points selected, leading to poor performance on downstream tasks. In this paper, we propose Keypoint Autoencoder, an unsupervised learning method for detecting keypoints. We encourage selecting sparse semantic keypoints by enforcing the reconstruction from keypoints to the original point cloud. To make sparse keypoint selection differentiable, Soft Keypoint Proposal is adopted by calculating weighted averages among input points. A downstream task of classifying shape with sparse keypoints is conducted to demonstrate the distinctiveness of our selected keypoints. Semantic Accuracy and Semantic Richness are proposed and our method gives competitive or even better performance than state of the arts on these two metrics.

cs.CV↗

arXiv2020

Influence of process conditions and pore morphology on the closure rate of pores in hot rolling of steel

Steel sheets are manufactured from slabs produced in continuous casting, which inevitably results in a porous initial microstructure. These pores are nuclei for ductile damage and need to be closed during rolling in regard of the strict performance requirements of todays advanced high strength steels. Due to the beneficial shape factor of the roll gap, recrystallization and high diffusion rates pore closure and elimination occurs primarily during hot rolling. In this paper, relevant influencing factors on pore closure are investigated based on representative volume elements (RVE). First, the load regime of typical multi-pass hot rolling schedules is calculated via macro scale FEM and then applied to micro scale RVEs. Within the RVE, shape changes of a given pore are tracked and the closure ratio is calculated. This paper elucidates the effect of different roll diameters, absolute reductions per pass, pore shape and pore orientation on the closure rate. The results show that the majority of pores can be closed during the initial hot rolling stages. However, pore closure can be accelerated through suitable process parameters. Furthermore, shape and orientation of pores should be taken into account, if a precise knowledge of the total height reduction, leading to complete pore closure, is desired.

physics.app-ph↗

arXiv2019

Attribute Guided Unpaired Image-to-Image Translation with Semi-supervised Learning

Unpaired Image-to-Image Translation (UIT) focuses on translating images among different domains by using unpaired data, which has received increasing research focus due to its practical usage. However, existing UIT schemes defect in the need of supervised training, as well as the lack of encoding domain information. In this paper, we propose an Attribute Guided UIT model termed AGUIT to tackle these two challenges. AGUIT considers multi-modal and multi-domain tasks of UIT jointly with a novel semi-supervised setting, which also merits in representation disentanglement and fine control of outputs. Especially, AGUIT benefits from two-fold: (1) It adopts a novel semi-supervised learning process by translating attributes of labeled data to unlabeled data, and then reconstructing the unlabeled data by a cycle consistency operation. (2) It decomposes image representation into domain-invariant content code and domain-specific style code. The redesigned style code embeds image style into two variables drawn from standard Gaussian distribution and the distribution of domain label, which facilitates the fine control of translation due to the continuity of both variables. Finally, we introduce a new challenge, i.e., disentangled transfer, for UIT models, which adopts the disentangled representation to translate data less related with the training set. Extensive experiments demonstrate the capacity of AGUIT over existing state-of-the-art models.

cs.CV↗

arXiv2014

Stochastic parallel gradient descent optimization based on decoupling of the software and hardware

We classified the decoupled stochastic parallel gradient descent (SPGD) optimization model into two different types: software and hardware decoupling methods. A kind of software decoupling method is then proposed and a kind of hardware decoupling method is also proposed depending on the Shack-Hartmann (S-H) sensor. Using the normal sensor to accelerate the convergence of algorithm, the hardware decoupling method seems a capable realization of decoupled method. Based on the numerical simulation for correction of phase distortion in atmospheric turbulence, our methods are analyzed and compared with basic SPGD model and also other decoupling models, on the aspects of different spatial resolutions, mismatched control channels and noise. The results show that the phase distortion can be compensated after tens iterations with a strong capacity of noise tolerance in our model.

astro-ph.IM↗

arXiv2014

Experimental study on Modified Linear Quadratic Gaussian Control for Adaptive Optics

To achieve high resolution imaging the standard control algorithm used for classical adaptive optics (AO) is the simple but efficient proportional-integral (PI) controller. The goal is to minimize the root mean square (RMS) error of the residual wave front. However, with the PI controller one does not reach this minimum. A possibility to achieve is to use Linear Quadratic Gaussian Control (LQG). In practice, however this control algorithm still encounters one unexpected problem, leading to the divergence of control in AO. In this paper we propose a Modified LQG (MLQG) to solve this issue. The controller is analyzed explicitly. Test in the lab shows strong stability and high precision compared to the classical control.

astro-ph.IM↗