arXiv subjects

Yan Feng

Publications and source records attributed to Yan Feng.

At least 73 records · Page 4Linked to original sources

arXiv2023

Numerical coupling of aerosol emissions, dry removal, and turbulent mixing in the E3SM Atmosphere Model version 1 (EAMv1), part I: dust budget analyses and the impacts of a revised coupling scheme

An earlier study evaluating the dust life cycle in the Energy Exascale Earth System Model (E3SM) Atmosphere Model version 1 (EAMv1) has revealed that the simulated global mean dust lifetime is substantially shorter when higher vertical resolution is used, primarily due to significant strengthening of dust dry removal in source regions. This paper demonstrates that the sequential splitting of aerosol emissions, dry removal, and turbulent mixing in the model's time integration loop, especially the calculation of dry removal after surface emissions and before turbulent mixing, is the primary reason for the vertical resolution sensitivity reported in that earlier study. Based on this reasoning, we propose a simple revision to the numerical process coupling scheme, which moves the application of the surface emissions to after dry removal and before turbulent mixing. The revised scheme allows newly emitted particles to be transported aloft by turbulence before being removed from the atmosphere, and hence better resembles the dust life cycle in the real world. Sensitivity experiments are conducted and analyzed to evaluate the impact of the revised coupling on the simulated aerosol climatology in EAMv1.

physics.comp-ph↗

arXiv2023

Pedestrian wayfinding behavior in a multi-story building: a comprehensive modeling study featuring route choice, wayfinding performance, and observation behavior

This paper proposes a comprehensive approach for modeling pedestrian wayfinding behavior in complex buildings. This study employs two types of discrete choice models (i.e., MNL and PSL) featuring pedestrian route choice behavior, and three multivariate linear regression (MLR) models featuring the overall wayfinding performance and observation behavior (e.g., hesitation behavior and head rotation). Behavioral and questionnaire data featuring pedestrian wayfinding behavior and personal information were collected using a Virtual Reality experiment. Four wayfinding tasks were designed to determine how personal, infrastructure, and route characteristics affect indoor pedestrian wayfinding behavior on three levels, including route choice, wayfinding performance, and observation behavior. We find that pedestrian route choice behavior is primarily influenced by route characteristics, whereas wayfinding performance is also influenced by personal characteristics. Observation behavior is mainly influenced by task complexity, personal characteristics, and local properties of the routes that offer route information. To the best of our knowledge, this work represents the first attempt to investigate the impact of the same comprehensive set of variables on various metrics feature wayfinding behavior simultaneously.

cs.HC↗

arXiv2023

Observation of fast sound in two-dimensional dusty plasma liquids

Equilibrium molecular dynamics simulations are performed to study two-dimensional (2D) dusty plasma liquids. Based on the stochastic thermal motion of simulated particles, the longitudinal and transverse phonon spectra are calculated, and used to determine the corresponding dispersion relations. From there, the longitudinal and transverse sound speeds of 2D dusty plasma liquids are obtained. It is discovered that, for wavenumbers beyond the hydrodynamic regime, the longitudinal sound speed of a 2D dusty plasma liquid exceeds its adiabatic value, i.e., the so-called fast sound. This phenomenon appears at roughly the same length scale of the cutoff wavenumber for transverse waves, confirming its relation to the emergent solidity of liquids in the non-hydrodynamic regime. Using the thermodynamic and transport coefficients extracted from the previous studies, and relying on the Frenkel theory, the ratio of the longitudinal to the adiabatic sound speeds is derived analytically, providing the optimal conditions for fast sound, which are in quantitative agreement with the current simulation results.

physics.plasm-ph↗

arXiv2023

Accelerating Diffusion Sampling with Classifier-based Feature Distillation

Although diffusion model has shown great potential for generating higher quality images than GANs, slow sampling speed hinders its wide application in practice. Progressive distillation is thus proposed for fast sampling by progressively aligning output images of $N$-step teacher sampler with $N/2$-step student sampler. In this paper, we argue that this distillation-based accelerating method can be further improved, especially for few-step samplers, with our proposed \textbf{C}lassifier-based \textbf{F}eature \textbf{D}istillation (CFD). Instead of aligning output images, we distill teacher's sharpened feature distribution into the student with a dataset-independent classifier, making the student focus on those important features to improve performance. We also introduce a dataset-oriented loss to further optimize the model. Experiments on CIFAR-10 show the superiority of our method in achieving high quality and fast sampling. Code is provided at \url{https://github.com/zju-SWJ/RCFD}.

cs.CV↗

arXiv2023

Statistical treatment of convolutional neural network super-resolution of inland surface wind for subgrid-scale variability quantification

Machine learning models have been employed to perform either physics-free data-driven or hybrid dynamical downscaling of climate data. Most of these implementations operate over relatively small downscaling factors because of the challenge of recovering fine-scale information from coarse data. This limits their compatibility with many global climate model outputs, often available between $\sim$50--100 km resolution, to scales of interest such as cloud resolving or urban scales. This study systematically examines the capability of convolutional neural networks (CNNs) to downscale surface wind speed data over land surface from different coarse resolutions (25 km, 48 km, and 100 km resolution) to 3 km. For each downscaling factor, we consider three CNN configurations that generate super-resolved predictions of fine-scale wind speed, which take between 1 to 3 input fields: coarse wind speed, fine-scale topography, and diurnal cycle. In addition to fine-scale wind speeds, probability density function parameters are generated, through which sample wind speeds can be generated accounting for the intrinsic stochasticity of wind speed. For generalizability assessment, CNN models are tested on regions with different topography and climate that are unseen during training. The evaluation of super-resolved predictions focuses on subgrid-scale variability and the recovery of extremes. Models with coarse wind and fine topography as inputs exhibit the best performance compared with other model configurations, operating across the same downscaling factor. Our diurnal cycle encoding results in lower out-of-sample generalizability compared with other input configurations.

physics.ao-ph↗

arXiv2023

Robust Sequence Networked Submodular Maximization

In this paper, we study the \underline{R}obust \underline{o}ptimization for \underline{se}quence \underline{Net}worked \underline{s}ubmodular maximization (RoseNets) problem. We interweave the robust optimization with the sequence networked submodular maximization. The elements are connected by a directed acyclic graph and the objective function is not submodular on the elements but on the edges in the graph. Under such networked submodular scenario, the impact of removing an element from a sequence depends both on its position in the sequence and in the network. This makes the existing robust algorithms inapplicable. In this paper, we take the first step to study the RoseNets problem. We design a robust greedy algorithm, which is robust against the removal of an arbitrary subset of the selected elements. The approximation ratio of the algorithm depends both on the number of the removed elements and the network topology. We further conduct experiments on real applications of recommendation and link prediction. The experimental results demonstrate the effectiveness of the proposed algorithm.

cs.AI↗

arXiv2023

Revealing the supercritical dynamics of dusty plasmas and their liquid-like to gas-like dynamical crossover

Dusty plasmas represent a powerful playground to study the collective dynamics of strongly coupled systems with important interdisciplinary connections to condensed matter physics. Due to the pure Yukawa repulsive interaction between dust particles, dusty plasmas do not display a traditional liquid-vapor phase transition, perfectly matching the definition of a supercritical fluid. Using molecular dynamics simulations, we verify the supercritical nature of dusty plasmas and reveal the existence of a dynamical liquid-like to gas-like crossover which perfectly matches the salient features of the Frenkel line in classical supercritical fluids. We present several diagnostics to locate this dynamical crossover spanning from local atomic connectivity, shear relaxation dynamics, velocity autocorrelation function, heat capacity, and various transport properties. All these different criteria well agree with each other and are able to successfully locate the Frenkel line in both 2D and 3D dusty plasmas. In addition, we propose the unity ratio of the instantaneous transverse sound speed $C_T$ to the average particle speed $\bar{v}_{p}$, i.e., $C_T / \bar{v}_{p} = 1$, as a new diagnostic to identify this dynamical crossover. Finally, we observe an emergent degree of universality in the collective dynamics and transport properties of dusty plasmas as a function of the screening parameter and dimensionality of the system. Intriguingly, the temperature of the dynamical transition is independent of the dimensionality, and it is found to be always $20$ times of the corresponding melting point. Our results open a new path for the study of single particle and collective dynamics in plasmas and their interrelation with supercritical fluids in general.

physics.plasm-ph↗

arXiv2023

Generalizable Black-Box Adversarial Attack with Meta Learning

In the scenario of black-box adversarial attack, the target model's parameters are unknown, and the attacker aims to find a successful adversarial perturbation based on query feedback under a query budget. Due to the limited feedback information, existing query-based black-box attack methods often require many queries for attacking each benign example. To reduce query cost, we propose to utilize the feedback information across historical attacks, dubbed example-level adversarial transferability. Specifically, by treating the attack on each benign example as one task, we develop a meta-learning framework by training a meta-generator to produce perturbations conditioned on benign examples. When attacking a new benign example, the meta generator can be quickly fine-tuned based on the feedback information of the new task as well as a few historical attacks to produce effective perturbations. Moreover, since the meta-train procedure consumes many queries to learn a generalizable generator, we utilize model-level adversarial transferability to train the meta-generator on a white-box surrogate model, then transfer it to help the attack against the target model. The proposed framework with the two types of adversarial transferability can be naturally combined with any off-the-shelf query-based attack methods to boost their performance, which is verified by extensive experiments.

cs.LG↗

arXiv2022

Bidirectional flow of two-dimensional dusty plasma under asymmetric periodic substrates driven by unbiased external excitations

Collective transport properties of a one-dimensional asymmetric periodic substrate (1DAPS) modulated two-dimensional dusty plasma (2DDP) driven by an unbiased sinusoidal excitation force are investigated using Langevin dynamical simulations. It is discovered that, by changing the amplitude and frequency of the unbiased sinusoidal external excitations, as well as the depth of 1DPAS, both the direction and speed of the persistent particle flow can be adjusted, i.e., both the flow rectification and its reversal of the ratchet effect of the steady drift motion for particles are achieved using various excitations. For the studied 2DDP under the 1DAPS, when the amplitude of the excitation increases from zero, the magnitude of the overall drift velocity increases from zero to its first maximum in the easy direction of the 1DAPS, next decreases gradually back to zero, and then increases from zero to its second maximum in the hard direction of 1DAPS before finally gradually decaying. It is found that, as the frequency of the excitation and the depth of 1DAPS change, the maximum overall drift velocity also varies, and the corresponding magnitude of the excitation varies simultaneously. The observed ratchet effect in both the easy and hard directions of 1DAPS for 2DDP is attributed to the combination of the spatial symmetry breaking of 1DAPS and the inertial effects of particles, which is further confirmed by the three different presented diagnostics.

physics.plasm-ph↗

arXiv2022

Dynamical commensuration effect in a two-dimensional Yukawa solid modulated by periodic substrates

Transverse depinning dynamics of a periodic-square-substrate modulated two-dimensional dusty plasma solid driven by a constant force in the longitudinal direction are investigated using Langevin dynamical simulations. During the increase of the commensuration ratio (the number ratio of particles to substrate potential wells), the nonmonotonic variation trend of the critical transverse depinning force is observed. It is found that the local maxima and minima of the critical transverse depinning force just correspond to the dynamical commensurate and incommensurate conditions, respectively. The dynamical commensurate effect is also clearly exhibited from the stable one-dimensional channel particle trajectories and the highly ordered structure, however, both the particle trajectories and the structure are more disordered under the incommensurate conditions. The nonmonotonic variation of the critical transverse depinning force is attributed to the stability of the lattice structure under various commensuration ratios.

physics.plasm-ph↗

arXiv2022

The crossed-sine wavefront sensor: first tests and results

The crossed-sine wavefront sensor (WFS) is a pupil plane wavefront sensor that measures the first derivatives of the wavefront. It is made by three main components: a gradient transmission filter (GTF) built from a product of sine functions rotated by 45 degrees around the optical axis, a 2x2 mini-lens array (MLA) at the focus of the tested optical system and a detector array located on a plane conjugated to the pupil. The basic principle consists in acquiring four pupil images simultaneously, each image being observed from different points located behind the GTF. After the simulation work which demonstrated the wavefront reconstruction capability, we are now in the phase of implementation of the prototype in the lab. The crossed-sine WFS could achieve a simultaneous high spatial resolution at the pupil of the tested optics and absolute measurement accuracy comparable to that attained by laser-interferometers. In this paper we introduce seven customized phase masks and make measurements of them.First tests and resultsare demonstrated, based on which we explore the performance of our crossed-sine WFS and make comparisons with that of the laser-interferomete

physics.optics↗

arXiv2022

Multi-Scale Architectures Matter: On the Adversarial Robustness of Flow-based Lossless Compression

As a probabilistic modeling technique, the flow-based model has demonstrated remarkable potential in the field of lossless compression \cite{idf,idf++,lbb,ivpf,iflow},. Compared with other deep generative models (eg. Autoregressive, VAEs) \cite{bitswap,hilloc,pixelcnn++,pixelsnail} that explicitly model the data distribution probabilities, flow-based models perform better due to their excellent probability density estimation and satisfactory inference speed. In flow-based models, multi-scale architecture provides a shortcut from the shallow layer to the output layer, which significantly reduces the computational complexity and avoid performance degradation when adding more layers. This is essential for constructing an advanced flow-based learnable bijective mapping. Furthermore, the lightweight requirement of the model design in practical compression tasks suggests that flows with multi-scale architecture achieve the best trade-off between coding complexity and compression efficiency.

cs.CV↗

arXiv2022

Superlubric-pinned transition of a two-dimensional solid dusty plasma under a periodic triangular substrate

The superlubric-pinned transition in the depinning dynamics of a two-dimensional (2D) solid dusty plasma modulated by 2D triangular periodic substrates is investigated using Langevin dynamical simulations. When the lattice structure of the 2D solid dusty plasma perfectly matches the triangular substrate, two distinctive pinned and moving ordered states are observed, as the external uniform driving force gradually increases from zero. When there is a mismatch between the lattice structure and the triangular substrate, however, on shallow substrates, it is discovered that all of the particles can slide freely on the substrate even when the applied driving force is tiny. This is a typical example of superlubricity, which is caused by the competition between the substrate-particle and particle-particle interactions. If the substrate depth increases further, as the driving force increases from zero, there are three dynamical states consisting of the pinned, the disordered plastic flow, and the moving ordered. In an underdense system, where there are fewer particles than potential well minima, it is found that the occurrence of the three different dynamical states is controlled by the depth of the substrate, which is quantitatively characterized using the average mobility.

physics.plasm-ph↗

arXiv2022

Directional Locking in a 2D Yukawa Solid Modulated by a 2D Periodic Substrate

Directional depinning dynamics of a 2D dusty plasma solid modulated by a 2D square periodic substrate are investigated using Langevin dynamical simulations. We observe prominent directional locking effects when the direction of the external driving force is varied relative to the underlying square substrate. These locking steps appear when the direction of the driving force is close to the symmetry direction of the substrate, corresponding to the different dynamical flow patterns and the structures. In the conditions between the adjacent locking steps, moving ordered states are observed. Although the discontinuous transitions often occur between the locking steps and the non-locking portion, the continuous transitions are also found around the locking step associated with the disordered plastic flow close to its termini. Our results show that directional locking also occurs for underdamped systems, which could be tested experimentally in dusty plasmas modulated by 2D substrates.

physics.plasm-ph↗

arXiv2022

sqSGD: Locally Private and Communication Efficient Federated Learning

Federated learning (FL) is a technique that trains machine learning models from decentralized data sources. We study FL under local notions of privacy constraints, which provides strong protection against sensitive data disclosures via obfuscating the data before leaving the client. We identify two major concerns in designing practical privacy-preserving FL algorithms: communication efficiency and high-dimensional compatibility. We then develop a gradient-based learning algorithm called \emph{sqSGD} (selective quantized stochastic gradient descent) that addresses both concerns. The proposed algorithm is based on a novel privacy-preserving quantization scheme that uses a constant number of bits per dimension per client. Then we improve the base algorithm in three ways: first, we apply a gradient subsampling strategy that simultaneously offers better training performance and smaller communication costs under a fixed privacy budget. Secondly, we utilize randomized rotation as a preprocessing step to reduce quantization error. Thirdly, an adaptive gradient norm upper bound shrinkage strategy is adopted to improve accuracy and stabilize training. Finally, the practicality of the proposed framework is demonstrated on benchmark datasets. Experiment results show that sqSGD successfully learns large models like LeNet and ResNet with local privacy constraints. In addition, with fixed privacy and communication level, the performance of sqSGD significantly dominates that of various baseline algorithms.

cs.LG↗

arXiv2022

Knowledge Distillation with the Reused Teacher Classifier

Knowledge distillation aims to compress a powerful yet cumbersome teacher model into a lightweight student model without much sacrifice of performance. For this purpose, various approaches have been proposed over the past few years, generally with elaborately designed knowledge representations, which in turn increase the difficulty of model development and interpretation. In contrast, we empirically show that a simple knowledge distillation technique is enough to significantly narrow down the teacher-student performance gap. We directly reuse the discriminative classifier from the pre-trained teacher model for student inference and train a student encoder through feature alignment with a single $\ell_2$ loss. In this way, the student model is able to achieve exactly the same performance as the teacher model provided that their extracted features are perfectly aligned. An additional projector is developed to help the student encoder match with the teacher classifier, which renders our technique applicable to various teacher and student architectures. Extensive experiments demonstrate that our technique achieves state-of-the-art results at the modest cost of compression ratio due to the added projector.

cs.CV↗

arXiv2021

Spectral Variability Augmented Sparse Unmixing of Hyperspectral Images

Spectral unmixing (SU) expresses the mixed pixels existed in hyperspectral images as the product of endmember and abundance, which has been widely used in hyperspectral imagery analysis. However, the influence of light, acquisition conditions and the inherent properties of materials, results in that the identified endmembers can vary spectrally within a given image (construed as spectral variability). To address this issue, recent methods usually use a priori obtained spectral library to represent multiple characteristic spectra of the same object, but few of them extracted the spectral variability explicitly. In this paper, a spectral variability augmented sparse unmixing model (SVASU) is proposed, in which the spectral variability is extracted for the first time. The variable spectra are divided into two parts of intrinsic spectrum and spectral variability for spectral reconstruction, and modeled synchronously in the SU model adding the regular terms restricting the sparsity of abundance and the generalization of the variability coefficient. It is noted that the spectral variability library and the intrinsic spectral library are all constructed from the In-situ observed image. Experimental results over both synthetic and real-world data sets demonstrate that the augmented decomposition by spectral variability significantly improves the unmixing performance than the decomposition only by spectral library, as well as compared to state-of-the-art algorithms.

eess.IV↗

arXiv2021

Phonon spectra of a two-dimensional solid dusty plasma modified by two-dimensional periodic substrates

Phonon spectra of a two-dimensional (2D) solid dusty plasma modulated by 2D square and triangular periodic substrates are investigated using Langevin dynamical simulations. The commensurability ratio, i.e., the ratio of the number of particles to the number of potential well minima, is set to 1 or 2. The resulting phonon spectra show that propagation of waves is always suppressed due to the confinement of particles by the applied 2D periodic substrates. For a commensurability ratio of 1, the spectra indicate that all particles mainly oscillate at one specific frequency, corresponding to the harmonic oscillation frequency of one single particle inside one potential well. At a commensurability ratio of 2, the substrate allows two particles to sit inside the bottom of each potential well, and the resulting longitudinal and transverse spectra exhibit four branches in total. We find that the two moderate branches come from the harmonic oscillations of one single particle and two combined particles in the potential well. The other two branches correspond to the relative motion of the two-body structure in each potential well in the radial and azimuthal directions. The difference in the spectra between the square and triangular substrates is attributed to the anisotropy of the substrates and the resulting alignment directions of the two-body structure in each potential well.

physics.plasm-ph↗