arXiv subjects

Zhe Xu

Publications and source records attributed to Zhe Xu.

At least 145 records · Page 8Linked to original sources

arXiv2023

Approximating Discontinuous Nash Equilibrial Values of Two-Player General-Sum Differential Games

Finding Nash equilibrial policies for two-player differential games requires solving Hamilton-Jacobi-Isaacs (HJI) PDEs. Self-supervised learning has been used to approximate solutions of such PDEs while circumventing the curse of dimensionality. However, this method fails to learn discontinuous PDE solutions due to its sampling nature, leading to poor safety performance of the resulting controllers in robotics applications when player rewards are discontinuous. This paper investigates two potential solutions to this problem: a hybrid method that leverages both supervised Nash equilibria and the HJI PDE, and a value-hardening method where a sequence of HJIs are solved with a gradually hardening reward. We compare these solutions using the resulting generalization and safety performance in two vehicle interaction simulation studies with 5D and 9D state spaces, respectively. Results show that with informative supervision (e.g., collision and near-collision demonstrations) and the low cost of self-supervised learning, the hybrid method achieves better safety performance than the supervised, self-supervised, and value hardening approaches on equal computational budget. Value hardening fails to generalize in the higher-dimensional case without informative supervision. Lastly, we show that the neural activation function needs to be continuously differentiable for learning PDEs and its choice can be case dependent.

cs.LG↗

arXiv2023

Propagation of spin channel waves

A mutilated model is constructed to approximate the collision term of spin Boltzmann equation that incorporates newly appearing collisional invariants i.e, the total angular momentum. With recourse to degenerate perturbation theory, the dispersion relations of hydrodynamic modes are formulated, among which spin modes are responsible for spin equilibration. We find that the non-locality does not change the sound speed but slows down the propagation of spin channel waves. The damping rates of spin modes are close to those of spinless modes over a reasonable parameter value range. The results reveal that both spin and momentum should be treated simultaneously in a unified transport framework. In the nonrelativistic limit, the short-wavelength behavior for normal modes is also explored and there exists a critical point for every distinct discrete mode over which only quasiparticle modes contribute.

hep-ph↗

arXiv2023

Constrained Active Classification Using Partially Observable Markov Decision Processes

In this work, we study the problem of actively classifying the attributes of dynamical systems characterized as a finite set of Markov decision process (MDP) models. We are interested in finding strategies that actively interact with the dynamical system and observe its reactions so that the attribute of interest is classified efficiently with high confidence. We present a decision-theoretic framework based on partially observable Markov decision processes (POMDPs). The proposed framework relies on assigning a classification belief (a probability distribution) to the attributes of interest. Given an initial belief, a confidence level over which a classification decision can be made, a cost bound, safe belief sets, and a finite time horizon, we compute POMDP strategies leading to classification decisions. We present three different algorithms to compute such strategies. The first algorithm computes the optimal strategy exactly by value iteration. To overcome the computational complexity of computing the exact solutions, we propose a second algorithm based on adaptive sampling and a third based on a Monte Carlo tree search to approximate the optimal probability of reaching a classification decision. We illustrate the proposed methodology using examples from medical diagnosis, security surveillance, and wildlife classification.

eess.SY↗

arXiv2022

CTT-Net: A Multi-view Cross-token Transformer for Cataract Postoperative Visual Acuity Prediction

Surgery is the only viable treatment for cataract patients with visual acuity (VA) impairment. Clinically, to assess the necessity of cataract surgery, accurately predicting postoperative VA before surgery by analyzing multi-view optical coherence tomography (OCT) images is crucially needed. Unfortunately, due to complicated fundus conditions, determining postoperative VA remains difficult for medical experts. Deep learning methods for this problem were developed in recent years. Although effective, these methods still face several issues, such as not efficiently exploring potential relations between multi-view OCT images, neglecting the key role of clinical prior knowledge (e.g., preoperative VA value), and using only regression-based metrics which are lacking reference. In this paper, we propose a novel Cross-token Transformer Network (CTT-Net) for postoperative VA prediction by analyzing both the multi-view OCT images and preoperative VA. To effectively fuse multi-view features of OCT images, we develop cross-token attention that could restrict redundant/unnecessary attention flow. Further, we utilize the preoperative VA value to provide more information for postoperative VA prediction and facilitate fusion between views. Moreover, we design an auxiliary classification loss to improve model performance and assess VA recovery more sufficiently, avoiding the limitation by only using the regression metrics. To evaluate CTT-Net, we build a multi-view OCT image dataset collected from our collaborative hospital. A set of extensive experiments validate the effectiveness of our model compared to existing methods in various metrics. Code is available at: https://github.com/wjh892521292/Cataract OCT.

eess.IV↗

arXiv2022

Learning Temporal Logic Properties: an Overview of Two Recent Methods

Learning linear temporal logic (LTL) formulas from examples labeled as positive or negative has found applications in inferring descriptions of system behavior. We summarize two methods to learn LTL formulas from examples in two different problem settings. The first method assumes noise in the labeling of the examples. For that, they define the problem of inferring an LTL formula that must be consistent with most but not all of the examples. The second method considers the other problem of inferring meaningful LTL formulas in the case where only positive examples are given. Hence, the first method addresses the robustness to noise, and the second method addresses the balance between conciseness and specificity (i.e., language minimality) of the inferred formula. The summarized methods propose different algorithms to solve the aforementioned problems, as well as to infer other descriptions of temporal properties, such as signal temporal logic or deterministic finite automata.

cs.LO↗

arXiv2022

Data Augmentation for Deep Graph Learning: A Survey

Graph neural networks, a powerful deep learning tool to model graph-structured data, have demonstrated remarkable performance on numerous graph learning tasks. To address the data noise and data scarcity issues in deep graph learning, the research on graph data augmentation has intensified lately. However, conventional data augmentation methods can hardly handle graph-structured data which is defined in non-Euclidean space with multi-modality. In this survey, we formally formulate the problem of graph data augmentation and further review the representative techniques and their applications in different deep graph learning problems. Specifically, we first propose a taxonomy for graph data augmentation techniques and then provide a structured review by categorizing the related work based on the augmented information modalities. Moreover, we summarize the applications of graph data augmentation in two representative problems in data-centric deep graph learning: (1) reliable graph learning which focuses on enhancing the utility of input graph as well as the model capacity via graph data augmentation; and (2) low-resource graph learning which targets on enlarging the labeled training data scale through graph data augmentation. For each problem, we also provide a hierarchical problem taxonomy and review the existing literature related to graph data augmentation. Finally, we point out promising research directions and the challenges in future research.

cs.LG↗

arXiv2022

Learn2Reg: comprehensive multi-task medical image registration challenge, dataset and evaluation in the era of deep learning

Image registration is a fundamental medical image analysis task, and a wide variety of approaches have been proposed. However, only a few studies have comprehensively compared medical image registration approaches on a wide range of clinically relevant tasks. This limits the development of registration methods, the adoption of research advances into practice, and a fair benchmark across competing approaches. The Learn2Reg challenge addresses these limitations by providing a multi-task medical image registration data set for comprehensive characterisation of deformable registration algorithms. A continuous evaluation will be possible at https://learn2reg.grand-challenge.org. Learn2Reg covers a wide range of anatomies (brain, abdomen, and thorax), modalities (ultrasound, CT, MR), availability of annotations, as well as intra- and inter-patient registration evaluation. We established an easily accessible framework for training and validation of 3D registration methods, which enabled the compilation of results of over 65 individual method submissions from more than 20 unique teams. We used a complementary set of metrics, including robustness, accuracy, plausibility, and runtime, enabling unique insight into the current state-of-the-art of medical image registration. This paper describes datasets, tasks, evaluation methods and results of the challenge, as well as results of further analysis of transferability to new datasets, the importance of label supervision, and resulting bias. While no single approach worked best across all tasks, many methodological aspects could be identified that push the performance of medical image registration to new state-of-the-art performance. Furthermore, we demystified the common belief that conventional registration methods have to be much slower than deep-learning-based methods.

eess.IV↗

arXiv2022

Distributed Differentially Private Control Synthesis for Multi-Agent Systems with Metric Temporal Logic Specifications

In this paper, we propose a distributed differentially private receding horizon control (RHC) approach for multi-agent systems (MAS) with metric temporal logic (MTL) specifications. In the MAS considered in this paper, each agent privatizes its sensitive information from other agents using a differential privacy mechanism. In other words, each agent adds privacy noise (e.g., Gaussian noise) to its output to maintain its privacy and communicates its noisy output with its neighboring agents. We define two types of MTL specifications for the MAS: agent-level specifications and system-level specifications. Agents should collaborate to satisfy the system-level MTL specifications with a minimum probability while each agent must satisfy its own agent-level MTL specifications at the same time. In the proposed distributed RHC approach, each agent communicates with its neighboring agents to acquire their noisy outputs and calculates an estimate of the system-level trajectory. Then each agent synthesizes its own control inputs such that the system-level specifications are satisfied with a minimum probability while the agent-level specifications are also satisfied. In the proposed optimization formulation of RHC, we directly incorporate Kalman filter equations to calculate the estimates of the system-level trajectory, and we use mixed-integer linear programming (MILP) to encode the MTL specifications as optimization constraints. Finally, we implement the proposed distributed RHC approach in a case study.

eess.SY↗

arXiv2022

A Hybrid Tracking Control Strategy for an Unmanned Underwater Vehicle Aided with Bioinspired Neural Dynamics

Tracking control has been a vital research topic in robotics. This paper presents a novel hybrid control strategy for an unmanned underwater vehicle (UUV) based on a bioinspired neural dynamics model. An enhanced backstepping kinematic control strategy is first developed to avoid sharp velocity jumps and provides smooth velocity commands relative to conventional methods. Then, a novel sliding mode control is proposed, which is capable of providing smooth and continuous torque commands free from chattering. In comparative studies, the proposed combined hybrid control strategy has ensured control signals smoothness, which is critical in real world applications, especially for an unmanned underwater vehicle that needs to operate in complex underwater environments.

cs.RO↗

arXiv2022

Non-Parametric Neuro-Adaptive Formation Control

We develop a learning-based algorithm for the distributed formation control of networked multi-agent systems governed by unknown, nonlinear dynamics. Most existing algorithms either assume certain parametric forms for the unknown dynamic terms or resort to unnecessarily large control inputs in order to provide theoretical guarantees. The proposed algorithm avoids these drawbacks by integrating neural network-based learning with adaptive control in a two-step procedure. In the first step of the algorithm, each agent learns a controller, represented as a neural network, using training data that correspond to a collection of formation tasks and agent parameters. These parameters and tasks are derived by varying the nominal agent parameters and a user-defined formation task to be achieved, respectively. In the second step of the algorithm, each agent incorporates the trained neural network into an online and adaptive control policy in such a way that the behavior of the multi-agent closed-loop system satisfies the user-defined formation task. Both the learning phase and the adaptive control policy are distributed, in the sense that each agent computes its own actions using only local information from its neighboring agents. The proposed algorithm does not use any a priori information on the agents' unknown dynamic terms or any approximation schemes. We provide formal theoretical guarantees on the achievement of the formation task.

eess.SY↗

arXiv2022

Towards Better Dermoscopic Image Feature Representation Learning for Melanoma Classification

Deep learning-based melanoma classification with dermoscopic images has recently shown great potential in automatic early-stage melanoma diagnosis. However, limited by the significant data imbalance and obvious extraneous artifacts, i.e., the hair and ruler markings, discriminative feature extraction from dermoscopic images is very challenging. In this study, we seek to resolve these problems respectively towards better representation learning for lesion features. Specifically, a GAN-based data augmentation (GDA) strategy is adapted to generate synthetic melanoma-positive images, in conjunction with the proposed implicit hair denoising (IHD) strategy. Wherein the hair-related representations are implicitly disentangled via an auxiliary classifier network and reversely sent to the melanoma-feature extraction backbone for better melanoma-specific representation learning. Furthermore, to train the IHD module, the hair noises are additionally labeled on the ISIC2020 dataset, making it the first large-scale dermoscopic dataset with annotation of hair-like artifacts. Extensive experiments demonstrate the superiority of the proposed framework as well as the effectiveness of each component. The improved dataset publicly avaliable at https://github.com/kirtsy/DermoscopicDataset.

eess.IV↗

arXiv2022

Reconfiguration and eruption of a solar filament by magnetic reconnection with an emerging magnetic field

Both observations and simulations suggest that the solar filament eruption is closely related to magnetic flux emergence. It is thought that the eruption is triggered by magnetic reconnection between the filament and the emerging flux. However, the details of such a reconnection are rarely presented. In this study, we report the detailed reconnection between a filament and its nearby emerging fields, that led to the reconfiguration and subsequent partial eruption of the filament located over the polarity inversion line of active region 12816. Before the reconnection, we observed repeated brightenings in the filament at a location that overlies a site of magnetic flux cancellation. Plasmoids form at this brightening region, and propagate bi-directionally along the filament. These indicate the tether-cutting reconnection that results in the formation and eruption of a flux rope. To the northwest of the filament, magnetic fields emerge, and reconnect with the context ones, resulting in repeated jets. Afterwards, another magnetic fields emerge near the northwestern filament endpoints, and reconnect with the filament, forming the newly reconnected filament and loops. Current sheet repeatedly occurs at the interface, with the mean temperature and emission measure of 1.7 MK and 1.1$\times$10$^{28}$ cm$^{-5}$. Plasmoids form in the current sheet, and propagate along it and further along the newly reconnected filament and loops. The newly reconnected filament then erupts, while the unreconnected filament remains stable. We propose that besides the orientation of emerging fields, some other parameters, such as the position, distance, strength, and area, are also crucial for triggering the filament eruption.

astro-ph.SR↗

arXiv2022

Deformer: Towards Displacement Field Learning for Unsupervised Medical Image Registration

Recently, deep-learning-based approaches have been widely studied for deformable image registration task. However, most efforts directly map the composite image representation to spatial transformation through the convolutional neural network, ignoring its limited ability to capture spatial correspondence. On the other hand, Transformer can better characterize the spatial relationship with attention mechanism, its long-range dependency may be harmful to the registration task, where voxels with too large distances are unlikely to be corresponding pairs. In this study, we propose a novel Deformer module along with a multi-scale framework for the deformable image registration task. The Deformer module is designed to facilitate the mapping from image representation to spatial transformation by formulating the displacement vector prediction as the weighted summation of several bases. With the multi-scale framework to predict the displacement fields in a coarse-to-fine manner, superior performance can be achieved compared with traditional and learning-based approaches. Comprehensive experiments on two public datasets are conducted to demonstrate the effectiveness of the proposed Deformer module as well as the multi-scale framework.

eess.IV↗

arXiv2022

Sunspot shearing and sudden retraction motion associated with the 2013 August 17 M3.3 Flare

In this Letter, we give a detailed analysis to the M3.3 class flare that occurred on August 17, 2013 (SOL2013-08-17T18:16). It presents a clear picture of mutual magnetic interaction initially from the photosphere to the corona via the abrupt rapid shearing motion of a small sunspot before the flare, and then suddenly from the corona back to the photosphere via the sudden retraction motion of the same sunspot during the flare impulsive phase. About 10 hours before the flare, a small sunspot in the active region NOAA 11818 started to move northeast along a magnetic polarity inversion line (PIL), creating a shearing motion that changed the quasi-static state of the active region. A filament right above the PIL was activated following the movement of the sunspot and then got partially erupted. The eruption eventually led to the M3.3 flare. The sunspot was then suddenly pulled back to the opposite direction upon the flare onset. During the backward motion, the Lorentz force underwent a simultaneous impulsive change both in magnitude and direction. Its directional change is found to be conformable with the retraction motion. The observation provides direct evidence for the role of the shearing motion of the sunspot in powering and triggering the flare. It especially confirms that the abrupt motion of a sunspot during a solar flare is the result of a back reaction caused by the reconfiguration of the coronal magnetic field.

astro-ph.SR↗

arXiv2022

Bio-inspired Intelligence with Applications to Robotics: A Survey

In the past decades, considerable attention has been paid to bio-inspired intelligence and its applications to robotics. This paper provides a comprehensive survey of bio-inspired intelligence, with a focus on neurodynamics approaches, to various robotic applications, particularly to path planning and control of autonomous robotic systems. Firstly, the bio-inspired shunting model and its variants (additive model and gated dipole model) are introduced, and their main characteristics are given in detail. Then, two main neurodynamics applications to real-time path planning and control of various robotic systems are reviewed. A bio-inspired neural network framework, in which neurons are characterized by the neurodynamics models, is discussed for mobile robots, cleaning robots, and underwater robots. The bio-inspired neural network has been widely used in real-time collision-free navigation and cooperation without any learning procedures, global cost functions, and prior knowledge of the dynamic environment. In addition, bio-inspired backstepping controllers for various robotic systems, which are able to eliminate the speed jump when a large initial tracking error occurs, are further discussed. Finally, the current challenges and future research directions are discussed in this paper.

cs.RO↗

arXiv2022

Seeking Common Ground While Reserving Differences: Multiple Anatomy Collaborative Framework for Undersampled MRI Reconstruction

Recently, deep neural networks have greatly advanced undersampled Magnetic Resonance Image (MRI) reconstruction, wherein most studies follow the one-anatomy-one-network fashion, i.e., each expert network is trained and evaluated for a specific anatomy. Apart from inefficiency in training multiple independent models, such convention ignores the shared de-aliasing knowledge across various anatomies which can benefit each other. To explore the shared knowledge, one naive way is to combine all the data from various anatomies to train an all-round network. Unfortunately, despite the existence of the shared de-aliasing knowledge, we reveal that the exclusive knowledge across different anatomies can deteriorate specific reconstruction targets, yielding overall performance degradation. Observing this, in this study, we present a novel deep MRI reconstruction framework with both anatomy-shared and anatomy-specific parameterized learners, aiming to "seek common ground while reserving differences" across different anatomies.Particularly, the primary anatomy-shared learners are exposed to different anatomies to model flourishing shared knowledge, while the efficient anatomy-specific learners are trained with their target anatomy for exclusive knowledge. Four different implementations of anatomy-specific learners are presented and explored on the top of our framework in two MRI reconstruction networks. Comprehensive experiments on brain, knee and cardiac MRI datasets demonstrate that three of these learners are able to enhance reconstruction performance via multiple anatomy collaborative learning.

eess.IV↗

arXiv2022

Consensus Formation Tracking for Multiple AUV Systems Using Distributed Bioinspired Sliding Mode Control

Consensus formation tracking of multiple autonomous underwater vehicles (AUVs) subject to nonlinear and uncertain dynamics is a challenging problem in robotics. To tackle this challenge, a distributed bioinspired sliding mode controller is proposed in this paper. First, the conventional sliding mode controller (SMC) is presented, and the consensus problem is addressed on the basis of graph theory. Next, to tackle the high frequency chattering issue in SMC scheme and meanwhile improve the robustness to the noises, a bioinspired approach is introduced, in which a neural dynamic model is employed to replace the nonlinear sign or saturation function in the synthesis of conventional sliding mode controllers. Furthermore, the input-to-state stability of the resulting closed-loop system is proved in the presence of bounded lumped disturbance by the Lyapunov stability theory. Finally, simulation experiments are conducted to demonstrate the effectiveness of the proposed distributed formation control protocol.

eess.SY↗

arXiv2022

Incomplete electromagnetic response of hot QCD matter

The electromagnetic response of hot QCD matter to decaying external magnetic fields is investigated. We examine the validity of Ohm's law and find that the induced electric current increases from zero and relaxes towards the value from Ohm's law. The relaxation time is larger than the lifetime of the external magnetic field for the QCD matter in relativistic heavy-ion collisions. The lower than expected electric current significantly suppresses the induced magnetic field and makes the electromagnetic response incomplete. We demonstrate the incomplete electromagnetic response of hot QCD matter by calculations employing the parton transport model combined with the solution of Maxwell's equations. Our results show a strong suppression by two orders of magnitude in the magnetic field, relatively to calculations assuming the validity of Ohm's law. This may undermine experimental efforts to measure magnetic-field-related effects in heavy-ion collisions.

hep-ph↗