arXiv subjects

Liang Chen

Publications and source records attributed to Liang Chen.

At least 325 records · Page 18Linked to original sources

arXiv2023

Cross-View Hierarchy Network for Stereo Image Super-Resolution

Stereo image super-resolution aims to improve the quality of high-resolution stereo image pairs by exploiting complementary information across views. To attain superior performance, many methods have prioritized designing complex modules to fuse similar information across views, yet overlooking the importance of intra-view information for high-resolution reconstruction. It also leads to problems of wrong texture in recovered images. To address this issue, we explore the interdependencies between various hierarchies from intra-view and propose a novel method, named Cross-View-Hierarchy Network for Stereo Image Super-Resolution (CVHSSR). Specifically, we design a cross-hierarchy information mining block (CHIMB) that leverages channel attention and large kernel convolution attention to extract both global and local features from the intra-view, enabling the efficient restoration of accurate texture details. Additionally, a cross-view interaction module (CVIM) is proposed to fuse similar features from different views by utilizing cross-view attention mechanisms, effectively adapting to the binocular scene. Extensive experiments demonstrate the effectiveness of our method. CVHSSR achieves the best stereo image super-resolution performance than other state-of-the-art methods while using fewer parameters. The source code and pre-trained models are available at https://github.com/AlexZou14/CVHSSR.

cs.CV↗

arXiv2023

MR Elastography with Optimization-Based Phase Unwrapping and Traveling Wave Expansion-based Neural Network (TWENN)

Magnetic Resonance Elastography (MRE) can characterize biomechanical properties of soft tissue for disease diagnosis and treatment planning. However, complicated wavefields acquired from MRE coupled with noise pose challenges for accurate displacement extraction and modulus estimation. Here we propose a pipeline for processing MRE images using optimization-based displacement extraction and Traveling Wave Expansion-based Neural Network (TWENN) modulus estimation. Phase unwrapping and displacement extraction were achieved by optimization of an objective function with Dual Data Consistency (Dual-DC). A complex-valued neural network using displacement covariance as input has been constructed for the estimation of complex wavenumbers. A model of traveling wave expansion is used to generate training datasets with different levels of noise for the network. The complex shear modulus map is obtained by a fusion of multifrequency and multidirectional data. Validation using images of brain and liver simulation demonstrates the practical value of the proposed pipeline, which can estimate the biomechanical properties with minimum root-mean-square-errors compared with state-of-the-art methods. Applications of the proposed method for processing MRE images of phantom, brain, and liver show clear anatomical features and that the pipeline is robust to noise and has a good generalization capability.

eess.IV↗

arXiv2023

Radio emission of tidal disruption events from wind-cloud interaction

Winds can be launched in tidal disruption event (TDE). It has been proposed that the winds can interact with the cloud surrounding the black hole, produce bow shocks, accelerate electrons, and produce radio emission. We restudy the wind-cloud interaction model. We employ the properties of winds found by the radiation hydrodynamic simulations of super-Eddington circularized accretion flow in TDEs. We can calculate the peak radio emission frequency, the luminosity at the peak frequency, and their time-evolution based on the TDEs wind-cloud interaction model. We find that the model predicted peak radio emission frequency, the luminosity at peak frequency, and their time evolution can be well consistent with those in TDEs AT2019dsg and ASASSN-14li. This indicates that in these two radio TDEs, the wind-cloud interaction mechanism may be responsible for the radio emission.

astro-ph.HE↗

arXiv2023

Neighborhood Consensus Contrastive Learning for Backward-Compatible Representation

In object re-identification (ReID), the development of deep learning techniques often involves model updates and deployment. It is unbearable to re-embedding and re-index with the system suspended when deploying new models. Therefore, backward-compatible representation is proposed to enable "new" features to be compared with "old" features directly, which means that the database is active when there are both "new" and "old" features in it. Thus we can scroll-refresh the database or even do nothing on the database to update. The existing backward-compatible methods either require a strong overlap between old and new training data or simply conduct constraints at the instance level. Thus they are difficult in handling complicated cluster structures and are limited in eliminating the impact of outliers in old embeddings, resulting in a risk of damaging the discriminative capability of new features. In this work, we propose a Neighborhood Consensus Contrastive Learning (NCCL) method. With no assumptions about the new training data, we estimate the sub-cluster structures of old embeddings. A new embedding is constrained with multiple old embeddings in both embedding space and discrimination space at the sub-class level. The effect of outliers diminished, as the multiple samples serve as "mean teachers". Besides, we also propose a scheme to filter the old embeddings with low credibility, further improving the compatibility robustness. Our method ensures backward compatibility without impairing the accuracy of the new model. And it can even improve the new model's accuracy in most scenarios.

cs.CV↗

arXiv2023

Locally Spontaneous Dynamic Oxygen Migration on Biphenylene: A DFT Study

The dynamic oxygen migration at the interface of carbon allotropes dominated by the periodic hexagonal rings, including graphene and carbon nanotube, has opened up a new avenue to realize dynamic covalent materials. However, for the carbon materials with hybrid carbon rings, such as biphenylene, whether the dynamic oxygen migration at its interface can still be found remains unknown. Using both density functional theory calculations and machine-learning-based molecular dynamics (MLMD) simulations, we found that the oxygen migration departing away from the four-membered carbon (C4) ring is hindered, and the oxygen atom prefers to spontaneously migrate toward/around the C4 ring. This locally spontaneous dynamic oxygen migration on the biphenylene is attributed to the high barrier of about 1.5 eV for the former process and relatively low barrier of about 0.3 eV for the latter one, originating from the enhanced activity of C-O bond near/around the C4 ring due to the hybrid carbon rings structure. Moreover, the locally spontaneous dynamic oxygen migration is further confirmed by MLMD simulations. This work sheds light on the potential of biphenylene as a catalyst for spatial-controlled energy conversion and provides the guidance for realizing the dynamic covalent interface at other carbon-based or two-dimensional materials.

cond-mat.mtrl-sci↗

arXiv2023

Solution of Volume Integral and Hydrodynamic Equations to Analyze Electromagnetic Scattering from Composite Nanostructures

A coupled system of volume integral and hydrodynamic equations is solved to analyze electromagnetic scattering from nanostructures consisting of metallic and dielectric parts. In the metallic part, the hydrodynamic equation relates the free electron polarization current to the electric flux and effectively "updates" the constitutive relation to enable the modeling of nonlocality. In the metallic and the dielectric parts, the volume integral equation relates the electric flux and the free electron polarization current to the scattered electric field. Unknown electric flux and free electron polarization current are expanded using Schaubert-Wilton-Glisson basis functions. Inserting these expansions into the coupled system of the volume integral and hydrodynamic equations and using Galerkin testing yield a matrix system in unknown expansion coefficients. An efficient two-level iterative solver is proposed to solve this matrix system. This approach "inverts" the discretized hydrodynamic equation for the coefficients of the free electron polarization current and substitutes the result in the discretized volume integral equation. Outer iterations solve this reduced matrix system while the inner iterations invert the discretized hydrodynamic equation at every iteration of the outer iterations. Numerical experiments are carried out to demonstrate the accuracy, the efficiency, and the applicability of the proposed method.

physics.comp-ph↗

arXiv2023

Tunable correlation effects of magnetic impurities by the cubic Rashba spin-orbit couplings

We theoretically study the influence of the $k$-cubic Rashba spin-orbit coupling (SOC) on the correlation effects of magnetic impurities by combining the variational method and the Hirsch-Fye quantum Monte Carlo (HFQMC) simulations. Markedly different from the normal $k$-linear Rashba SOC, even a small cubic Rashba term can greatly alter the band structure and induce a Van Hove singularity in a wide range of energy, thus the single impurity local moment becomes largely tunable. The cubic Rashba SOC adopted in this work breaks the rotational symmetry, but the host material is still invariant under the operations $\mathcal{R}^z(π)$, $\mathcal{IR}^z(π/2)$, $\mathcal{M}_{xz}$, $\mathcal{M}_{yz}$, where $\mathcal{R}^z(θ)$ is the rotation of angle $θ$ about the $z$-axis, $\mathcal{I}$ is the inversion operator and $\mathcal{M}_{xz}$ ($\mathcal{M}_{yz}$) is the mirror reflection about the $x$-$z$ ($y$-$z$) principal plane. Saliently, various components of spin-spin correlation between the single magnetic impurity and the conduction electrons show three- or six-fold rotational symmetry. This unique feature is due to the triple winding of the spins with a $2π$ rotation of $\mathbf{k}$, which is a hallmark of the cubic Rashba effect and can possibly be an identifier to distinguish the cubic Rashba SOC from the normal $k$-linear Rashba term in experiments. Although the cubic Rashba term drastically alters the electronic properties of the host, we find that the spatial decay rate of the spin-spin correlation function remains essentially unchanged. Moreover, the carrier-mediated Ruderman-Kittel-Kasuya-Yosida interactions between two magnetic impurities show twisted features, the ferromagnetic diagonal terms dominate when two magnetic impurities are very close, but the off-diagonal terms become important at long distances.

cond-mat.str-el↗

arXiv2023

Self-Sampling Training and Evaluation for the Accuracy-Bias Tradeoff in Recommendation

Research on debiased recommendation has shown promising results. However, some issues still need to be handled for its application in industrial recommendation. For example, most of the existing methods require some specific data, architectures and training methods. In this paper, we first argue through an online study that arbitrarily removing all the biases in industrial recommendation may not consistently yield a desired performance improvement. For the situation that a randomized dataset is not available, we propose a novel self-sampling training and evaluation (SSTE) framework to achieve the accuracy-bias tradeoff in recommendation, i.e., eliminate the harmful biases and preserve the beneficial ones. Specifically, SSTE uses a self-sampling module to generate some subsets with different degrees of bias from the original training and validation data. A self-training module infers the beneficial biases and learns better tradeoff based on these subsets, and a self-evaluation module aims to use these subsets to construct more plausible references to reflect the optimized model. Finally, we conduct extensive offline experiments on two datasets to verify the effectiveness of our SSTE. Moreover, we deploy our SSTE in homepage recommendation of a famous financial management product called Tencent Licaitong, and find very promising results in an online A/B test.

cs.IR↗

arXiv2023

CARD: Semantic Segmentation with Efficient Class-Aware Regularized Decoder

Semantic segmentation has recently achieved notable advances by exploiting "class-level" contextual information during learning. However, these approaches simply concatenate class-level information to pixel features to boost the pixel representation learning, which cannot fully utilize intra-class and inter-class contextual information. Moreover, these approaches learn soft class centers based on coarse mask prediction, which is prone to error accumulation. To better exploit class level information, we propose a universal Class-Aware Regularization (CAR) approach to optimize the intra-class variance and inter-class distance during feature learning, motivated by the fact that humans can recognize an object by itself no matter which other objects it appears with. Moreover, we design a dedicated decoder for CAR (CARD), which consists of a novel spatial token mixer and an upsampling module, to maximize its gain for existing baselines while being highly efficient in terms of computational cost. Specifically, CAR consists of three novel loss functions. The first loss function encourages more compact class representations within each class, the second directly maximizes the distance between different class centers, and the third further pushes the distance between inter-class centers and pixels. Furthermore, the class center in our approach is directly generated from ground truth instead of from the error-prone coarse prediction. CAR can be directly applied to most existing segmentation models during training, and can largely improve their accuracy at no additional inference overhead. Extensive experiments and ablation studies conducted on multiple benchmark datasets demonstrate that the proposed CAR can boost the accuracy of all baseline models by up to 2.23% mIOU with superior generalization ability. CARD outperforms SOTA approaches on multiple benchmarks with a highly efficient architecture.

cs.CV↗

arXiv2022

A Cut-and-solve Algorithm for Virtual Machine Consolidation Problem

The virtual machine consolidation problem (VMCP) attempts to determine which servers to be activated, how to allocate virtual machines (VMs) to the activated servers, and how to migrate VMs among servers such that the summation of activated, allocation, and migration costs is minimized subject to the resource constraints of the servers and other practical constraints. In this paper, we first propose a new mixed integer linear programming (MILP) formulation for the VMCP. We show that compared with existing formulations, the proposed formulation is much more compact in terms of smaller numbers of variables or constraints, which makes it suitable for solving large-scale problems. We then develop a cut-and-solve (C&S) algorithm, a tree search algorithm to efficiently solve the VMCP to optimality. The proposed C&S algorithm is based on a novel relaxation of the VMCP that provides a stronger lower bound than the natural continuous relaxation of the VMCP, making a smaller search tree. By extensive computational experiments, we show that (i) the proposed formulation significantly outperforms existing formulations in terms of solution efficiency; and (ii) compared with standard MILP solvers, the proposed C&S algorithm is much more efficient.

cs.DS↗

arXiv2022

Flux Variations of Cosmic Ray Air Showers Detected by LHAASO-KM2A During a Thunderstorm on 10 June 2021

The Large High Altitude Air Shower Observatory (LHAASO) has three sub-arrays, KM2A, WCDA and WFCTA. The flux variations of cosmic ray air showers were studied by analyzing the KM2A data during the thunderstorm on 10 June 2021. The number of shower events that meet the trigger conditions increases significantly in atmospheric electric fields, with maximum fractional increase of 20%. The variations of trigger rates (increases or decreases) are found to be strongly dependent on the primary zenith angle. The flux of secondary particles increases significantly, following a similar trend with that of the shower events. To better understand the observed behavior, Monte Carlo simulations are performed with CORSIKA and G4KM2A (a code based on GEANT4). We find that the experimental data (in saturated negative fields) are in good agreement with simulations, assuming the presence of a uniform upward electric field of 700 V/cm with a thickness of 1500 m in the atmosphere above the observation level. Due to the acceleration/deceleration and deflection by the atmospheric electric field, the number of secondary particles with energy above the detector threshold is modified, resulting in the changes in shower detection rate.

astro-ph.HE↗

arXiv2022

Spectral Adversarial Training for Robust Graph Neural Network

Recent studies demonstrate that Graph Neural Networks (GNNs) are vulnerable to slight but adversarially designed perturbations, known as adversarial examples. To address this issue, robust training methods against adversarial examples have received considerable attention in the literature. \emph{Adversarial Training (AT)} is a successful approach to learning a robust model using adversarially perturbed training samples. Existing AT methods on GNNs typically construct adversarial perturbations in terms of graph structures or node features. However, they are less effective and fraught with challenges on graph data due to the discreteness of graph structure and the relationships between connected examples. In this work, we seek to address these challenges and propose Spectral Adversarial Training (SAT), a simple yet effective adversarial training approach for GNNs. SAT first adopts a low-rank approximation of the graph structure based on spectral decomposition, and then constructs adversarial perturbations in the spectral domain rather than directly manipulating the original graph structure. To investigate its effectiveness, we employ SAT on three widely used GNNs. Experimental results on four public graph datasets demonstrate that SAT significantly improves the robustness of GNNs against adversarial attacks without sacrificing classification accuracy and training efficiency.

cs.LG↗

arXiv2022

Are All Edges Necessary? A Unified Framework for Graph Purification

Graph Neural Networks (GNNs) as deep learning models working on graph-structure data have achieved advanced performance in many works. However, it has been proved repeatedly that, not all edges in a graph are necessary for the training of machine learning models. In other words, some of the connections between nodes may bring redundant or even misleading information to downstream tasks. In this paper, we try to provide a method to drop edges in order to purify the graph data from a new perspective. Specifically, it is a framework to purify graphs with the least loss of information, under which the core problems are how to better evaluate the edges and how to delete the relatively redundant edges with the least loss of information. To address the above two problems, we propose several measurements for the evaluation and different judges and filters for the edge deletion. We also introduce a residual-iteration strategy and a surrogate model for measurements requiring unknown information. The experimental results show that our proposed measurements for KL divergence with constraints to maintain the connectivity of the graph and delete edges in an iterative way can find out the most edges while keeping the performance of GNNs. What's more, further experiments show that this method also achieves the best defense performance against adversarial attacks.

cs.SI↗

arXiv2022

Competition-Induced Sign Reversal of Casimir-Lifshitz Torque: An Investigation on Topological Node-Line Semimetal

The dispersion of quasiparticles in topological node-line semimetals is significantly different in different directions. In a certain direction, the quasiparticles behave like relativistic particles with constant velocity. In other directions, they act as two-dimensional electron gas. The competition between relativistic and nonrelativistic dispersions can induce a sign reversal of Casimir-Lifshitz torque. Three different approaches can be applied to generate this sign reversal, i.e., tuning the anisotropic parameter or chemical potential in node-line semimetal, changing the distance between this material and substrate birefringence. Detailed calculations are illustrated for the system with topological node-line semimetal Ca$_3$P$_2$ and liquid crystal material 4-cyano-4-n-pentylcyclohexane-phenyl.

cond-mat.mes-hall↗

arXiv2022

Ultrafast ion sieving in two dimensional graphene oxide membranes

Ultrahigh water permeance, together with a high rejection rate through nanofiltration and separation membranes1,2, is crucial but still challenging for multivalent ion sieving in water treatment processes of desalination, separation, and purification3,4. To date, no theory or equation has ever been quantitatively clarified the mechanism of water permeance in two-dimensional (2D) membranes, despite intensive and prolonged searches. Here, we established a new general equation of permeation through 2D membranes, and experimentally achieved unprecedented advances in water permeance one to two orders of magnitude higher than state-of-the-art membranes while simultaneously maintaining high ion rejection rates for multivalent metal ions, by staking nano-sized reduced graphene oxide (nano-rGO) flakes into nanofiltration membranes. The equation is simply based on a fundamental steady-state flow assumption and provides an essential description of water permeance through 2D membranes, demonstrating that the ultrahigh water permeance is attributed to the high effective channel area and shortened channel length elicited from the nano-sized-flake stacking effects in nano-rGO membranes, consistent with our theoretical simulations and previous experiments. These results pave the way for fabrication of advanced 2D nanofiltration membranes to realize a breakthrough in water permeance with exceptional ion sieving performance.

physics.flu-dyn↗

arXiv2022

Large-batch Optimization for Dense Visual Predictions

Training a large-scale deep neural network in a large-scale dataset is challenging and time-consuming. The recent breakthrough of large-batch optimization is a promising way to tackle this challenge. However, although the current advanced algorithms such as LARS and LAMB succeed in classification models, the complicated pipelines of dense visual predictions such as object detection and segmentation still suffer from the heavy performance drop in the large-batch training regime. To address this challenge, we propose a simple yet effective algorithm, named Adaptive Gradient Variance Modulator (AGVM), which can train dense visual predictors with very large batch size, enabling several benefits more appealing than prior arts. Firstly, AGVM can align the gradient variances between different modules in the dense visual predictors, such as backbone, feature pyramid network (FPN), detection, and segmentation heads. We show that training with a large batch size can fail with the gradient variances misaligned among them, which is a phenomenon primarily overlooked in previous work. Secondly, AGVM is a plug-and-play module that generalizes well to many different architectures (e.g., CNNs and Transformers) and different tasks (e.g., object detection, instance segmentation, semantic segmentation, and panoptic segmentation). It is also compatible with different optimizers (e.g., SGD and AdamW). Thirdly, a theoretical analysis of AGVM is provided. Extensive experiments on the COCO and ADE20K datasets demonstrate the superiority of AGVM. For example, it can train Faster R-CNN+ResNet50 in 4 minutes without losing performance. AGVM enables training an object detector with one billion parameters in just 3.5 hours, reducing the training time by 20.9x, whilst achieving 62.2 mAP on COCO. The deliverables are released at https://github.com/Sense-X/AGVM.

cs.CV↗

arXiv2022

A Two-Stage Method for Chinese AMR Parsing

In this paper, we provide a detailed description of our system at CAMRP-2022 evaluation. We firstly propose a two-stage method to conduct Chinese AMR Parsing with alignment generation, which includes Concept-Prediction and Relation-Prediction stages. Our model achieves 0.7756 and 0.7074 Align-Smatch F1 scores on the CAMR 2.0 test set and the blind-test set of CAMRP-2022 individually. We also analyze the result and the limitation such as the error propagation and class imbalance problem we conclude in the current method. Code and the trained models are released at https://github.com/PKUnlp-icler/Two-Stage-CAMRP for reproduction.

cs.CL↗

arXiv2022

Uncertainty-Induced Transferability Representation for Source-Free Unsupervised Domain Adaptation

Source-free unsupervised domain adaptation (SFUDA) aims to learn a target domain model using unlabeled target data and the knowledge of a well-trained source domain model. Most previous SFUDA works focus on inferring semantics of target data based on the source knowledge. Without measuring the transferability of the source knowledge, these methods insufficiently exploit the source knowledge, and fail to identify the reliability of the inferred target semantics. However, existing transferability measurements require either source data or target labels, which are infeasible in SFUDA. To this end, firstly, we propose a novel Uncertainty-induced Transferability Representation (UTR), which leverages uncertainty as the tool to analyse the channel-wise transferability of the source encoder in the absence of the source data and target labels. The domain-level UTR unravels how transferable the encoder channels are to the target domain and the instance-level UTR characterizes the reliability of the inferred target semantics. Secondly, based on the UTR, we propose a novel Calibrated Adaption Framework (CAF) for SFUDA, including i)the source knowledge calibration module that guides the target model to learn the transferable source knowledge and discard the non-transferable one, and ii)the target semantics calibration module that calibrates the unreliable semantics. With the help of the calibrated source knowledge and the target semantics, the model adapts to the target domain safely and ultimately better. We verified the effectiveness of our method using experimental results and demonstrated that the proposed method achieves state-of-the-art performances on the three SFUDA benchmarks. Code is available at https://github.com/SPIresearch/UTR.

cs.CV↗