arXiv.itsnick.co

an independent field guide to 23 million papers from arXiv & OpenAlex — not affiliated with arXiv.org

Author

Siming He

52 papers · 2005–2026

2005: 12006: 02007: 02008: 02009: 02010: 02011: 02012: 12013: 02014: 02015: 02016: 12017: 32018: 02019: 52020: 22021: 82022: 22023: 52024: 82025: 72026: 9
2005peak 9 in 20262026

Frequent co-authors

Papers

showing 1–25 of 52 RSS

Are we really making much progress? Revisiting, benchmarking, and refining heterogeneous graph neural networks

Qingsong Lv, Ming Ding, Qiang Liu +7 more

cs.LG · · cited 515× · arXiv ↗ · more like this ·

Heterogeneous graph neural networks (HGNNs) have been blossoming in recent years, but the unique data processing and evaluation setups used by each work obstruct a full understanding of their advancements. In this work, we present a systematical reproduction of 12 recent HGNNs by using their official codes, datasets, settings, and hyperparameters, revealing surprising findings about the progress of HGNNs. We find that the simple homogeneous GNNs, e.g., GCN and GAT, are largely underestimated due to improper settings. GAT with proper inputs can generally match or outperform all existing HGNNs across various scenarios. To facilitate robust and reproducible HGNN research, we construct the Heterogeneous Graph Benchmark (HGB), consisting of 11 diverse datasets with three tasks. HGB standardizes the process of heterogeneous graph data splits, feature processing, and performance evaluation.…

Are we really making much progress?

Qingsong Lv, Ming Ding, Qiang Liu +7 more

Artificial Intelligence · · cited 253× · DOI ↗ · more like this ·

Heterogeneous graph neural networks (HGNNs) have been blossoming in recent years, but the unique data processing and evaluation setups used by each work obstruct a full understanding of their advancements. In this work, we present a systematical reproduction of 12 recent HGNNs by using their official codes, datasets, settings, and hyperparameters, revealing surprising findings about the progress of HGNNs. We find that the simple homogeneous GNNs, e.g., GCN and GAT, are largely underestimated due to improper settings. GAT with proper inputs can generally match or outperform all existing HGNNs across various scenarios. To facilitate robust and reproducible HGNN research, we construct the Heterogeneous Graph Benchmark (HGB) , consisting of 11 diverse datasets with three tasks. HGB standardizes the process of heterogeneous graph data splits, feature processing, and performance evaluation.…

Research of a Sign Language Translation System Based on Deep Learning

Siming He

Human-Computer Interaction · · cited 95× · DOI ↗ · more like this ·

Sign language is a visual language that uses hand gesture, change of hand shape and track information to express meaning, and is the main communication tool for people with hearing and language impairment. Sign language recognition can improve the problem that the number of people who need to use sign language is large but the popularity of sign language is poor, and provide a more convenient way of study, work and life for people with hearing and language impairment. Hand locating and sign language recognition methods can generally be divided into traditional methods and deep learning methods. In recent years, with the brilliant achievements of deep learning in the field of computer vision, it has been proved that the method based on deep learning has many advantages, such as rich feature extraction, strong modeling ability and intuitive training. Therefore, this paper studies hand…

Suppression of blow-up in Patlak-Keller-Segel via shear flows

Jacob Bedrossian, Siming He

math.AP · · cited 84× · arXiv ↗ · more like this ·

In this paper we consider the parabolic-elliptic Patlak-Keller-Segel models in $\mathbb T^d$ with $d=2,3$ with the additional effect of advection by a large shear flow. Without the shear flow, the model is $L^1$ critical in two dimensions with critical mass $8π$; solutions with mass less than $8π$ are global and solutions with mass larger than $8 π$ with finite second moment, all blow up in finite time. In three dimensions, the model is $L^{3/2}$ critical and $L^1$ supercritical; there exists solutions with arbitrarily small mass which blow up in finite time arbitrarily fast. We show that the additional shear flow, if it is chosen sufficiently large, suppresses one dimension of the dynamics and hence can suppress blow-up. In two dimensions, the problem becomes effectively $L^1$ subcritical and so all solutions are global in time (if the shear flow is chosen large). In three dimensions…

Suppression of blow-up in Parabolic-Parabolic Patlak-Keller-Segel via strictly monotone shear flows

Siming He

math.AP · · cited 67× · arXiv ↗ · more like this ·

In this paper we consider the parabolic-parabolic Patlak-Keller-Segel models in $\mathbb{T}\times\mathbb{R}$ with advection by a large strictly monotone shear flow. Without the shear flow, the model is $L^1$ critical in two dimensions with critical mass $8π$: solutions with mass less than $8π$ are global in time and there exist solutions with mass larger than $8 π$ which blow up in finite time \cite{Schweyer14}. We show that the additional shear flow, if it is chosen sufficiently large, suppresses one dimension of the dynamics and hence can suppress blow-up. In contrast with the parabolic-elliptic case \cite{BedrossianHe16}, the strong shear flow has destabilizing effect in addition to the enhanced dissipation effect, which make the problem more difficult.

Global regularity of two-dimensional flocking hydrodynamics

Siming He, Eitan Tadmor

math.AP · · cited 60× · arXiv ↗ · more like this ·

We study the systems of Euler equations which arise from agent-based dynamics driven by velocity \emph{alignment}. It is known that smooth solutions of such systems must flock, namely -- the large time behavior of the velocity field approaches a limiting "flocking" velocity. To address the question of global regularity, we derive sharp critical thresholds in the phase space of initial configuration which characterize the global regularity and hence flocking behavior of such \emph{two-dimensional} systems. Specifically, we prove for that a large class of \emph{sub-critical} initial conditions such that the initial divergence is "not too negative" and the initial spectral gap is "not too large", global regularity persists for all time.

On the $8\pi$-Critical-Mass Threshold of a Patlak--Keller--Segel--Navier--Stokes System

Yishu Gong, Siming He

Modeling and Simulation · SIAM Journal on Mathematical Analysis · · cited 26× · DOI ↗ · more like this ·

Related DatabasesWeb of Science You must be logged in with an active subscription to view this.Article DataHistorySubmitted: 26 May 2020Accepted: 17 February 2021Published online: 20 May 2021KeywordsPatlak--Keller--Segel--Navier--Stokes system, critical mass, long-time behaviorAMS Subject Headings35B40, 35B45Publication DataISSN (print): 0036-1410ISSN (online): 1095-7154Publisher: Society for Industrial and Applied MathematicsCODEN: sjmaah

A game of alignment:collective behavior of multi-species

Siming He, Eitan Tadmor

math.AP · · cited 20× · arXiv ↗ · more like this ·

We study the (hydro-)dynamics of multi-species driven by alignment. What distinguishes the different species is the protocol of their interaction with the rest of the crowd: the collective motion is described by different communication kernels, $ϕ_{αβ}$, between the crowds in species $α$ and $β$. We show that flocking of the overall crowd emerges provided the communication array between species forms a connected graph. In particular, the crowd within each species need not interact with its own kind, i.e., $ϕ_{αα}=0$; different species which are engaged in such `game' of alignment require a connecting path for propagation of information which will lead to the flocking of overall crowd. The same methodology applies to multi-species aggregation dynamics governed by first-order alignment: connectivity implies concentration around an emerging consensus.

Multi-species Patlak-Keller-Segel system

Siming He, Eitan Tadmor

math.AP · · cited 8× · arXiv ↗ · more like this ·

We study the regularity and large-time behavior of a crowd of species driven by chemo-tactic interactions. What distinguishes the different species is the way they interact with the rest of the crowd: the collective motion is driven by different chemical reactions which end up in a coupled system of parabolic Patlak-Keller-Segel equations. We show that the densities of the different species diffuse to zero provided the chemical interactions between the different species satisfy certain sub-critical condition; the latter is intimately related to a log-Hardy-Littlewood-Sobolev inequality for systems due to Shafrir & Wolansky. Thus for example, when two species interact, one of which has mass less than $4π$, then the 2-system stays smooth for all time independent of the total mass of the system, in sharp contrast with the well-known breakdown of one specie with initial mass$> 8π$.

On the Fast Spreading Scenario

Siming He, Eitan Tadmor, Andrej Zlatoš

math.AP · · cited 7× · arXiv ↗ · more like this ·

We study two types of divergence-free fluid flows on unbounded domains in two and three dimensions -- hyperbolic and shear flows -- and their influence on chemotaxis and combustion. We show that fast spreading by these flows, when they are strong enough, can suppress growth of solutions to PDE modeling these phenomena. This includes prevention of singularity formation and global regularity of solutions to advective Patlak-Keller-Segel equations on ${\mathbb R}^2$ and ${\mathbb R}^3$, confirming numerical observations in [Khan, Johnson,Cartee and Yao, Global regularity of chemotaxis equations with advection, Involve, 9(1) 2016], as well as quenching in advection-reaction-diffusion equations.

Effect Evaluation of Multilevel Fuzzy Comprehensive Evaluation Mechanism Combined with Multimedia Teaching on the Cultivation of Comprehensive Quality of Vocational Students

Jin Si, Siming He

Education · Advances in Multimedia · · cited 6× · DOI ↗ · more like this ·

The core goal of higher vocational education is to cultivate students with good professional quality. The employability of college students depends on their professional quality to a large extent. From the perspective of school education, all kinds of higher vocational colleges have entered a new level in practice, but the evaluation of comprehensive quality is still relatively weak. Therefore, in order to enhance the comprehensive ability of higher vocational students and cultivate their excellent comprehensive quality, this paper adopts the fuzzy comprehensive evaluation method and puts forward the evaluation method of the comprehensive evaluation system of higher vocational students, so as to provide reference for solving the problem of comprehensive quality evaluation. The specific process is as follows: in this paper, a two-level fuzzy mathematical model is established by using…

Active Perception using Neural Radiance Fields

Siming He, Christopher D. Hsu, Dexter Ong +2 more

cs.RO · · cited 5× · arXiv ↗ · more like this ·

We study active perception from first principles to argue that an autonomous agent performing active perception should maximize the mutual information that past observations posses about future ones. Doing so requires (a) a representation of the scene that summarizes past observations and the ability to update this representation to incorporate new observations (state estimation and mapping), (b) the ability to synthesize new observations of the scene (a generative model), and (c) the ability to select control trajectories that maximize predictive information (planning). This motivates a neural radiance field (NeRF)-like representation which captures photometric, geometric and semantic properties of the scene grounded. This representation is well-suited to synthesizing new observations from different viewpoints. And thereby, a sampling-based planner can be used to calculate the…

Distance comparison principle and Grayson type theorem in the three dimensional curve shortening flow

Siming He

math.DG · · cited 5× · arXiv ↗ · more like this ·

In this paper, we use the distance comparison principle, first been developed by G. Huisken, to study the spatial curve shortening flow. We have got the result that if the initial curve is the helix, then the local minimum of the ratio of the extrinsic and intrinsic distance is non-decreasing. And we have proved a Grayson type theorem for a variety of spatial curves.

Time-dependent shear flows and their applications in parabolic–parabolic Patlak–Keller–Segel systems<sup>*</sup>

Siming He

Modeling and Simulation · Nonlinearity · · cited 3× · DOI ↗ · more like this ·

Abstract In this study, we investigate the behavior of three-dimensional parabolic–parabolic Patlak–Keller–Segel systems in the presence of ambient shear flows. Our findings demonstrate that when the total mass of the cell density is below a specific threshold, the solution remains globally regular as long as the flow is sufficiently strong. The primary difficulty in our analysis stems from the fast creation of chemical gradients due to strong shear advection.

Morphometric Imitation: From Morphology and Contact Aware Hand Retargeting to Sim-to-Real Visuomotor Policy

Tara Sadjadpour, Siming He, C. K. Wolfe +5 more

cs.RO · · cited 2× · arXiv ↗ · more like this ·

Human hand-object interactions (HOIs) provide a rich source of demonstrations for dexterous manipulation, but learning directly from them presents challenges in bridging morphology gaps, ensuring dynamical feasibility, and sim-to-real deployment. We present Morphometric Imitation, a three-stage framework that transforms reconstructed HOIs into zero-shot sim-to-real visuomotor policies. First, morphometric optimization (MMO) kinematically retargets human motion across hand morphologies while preserving demonstrated contacts. Second, residual reinforcement learning (RL) refines the kinematic reference using object pose and contact information from the human motion to produce dynamically feasible robot demonstrations. Third, these demonstrations are distilled into visuomotor policies. Across three robot hands and ten HOIs, MMO improves contact F1 over the strongest of five baselines by at…

Riemann-Hilbert Problems for biaxial symmetric solutions to iterated Dirac equations with <i>α</i> -weight

Siming He, Chunxue Duan, Yonghong Xie

Applied Mathematics · Applicable Analysis · · cited 2× · DOI ↗ · more like this ·

In this paper, Riemann-Hilbert boundary value problems (RHBVPs) with variable coefficients for biaxial symmetric solutions to iterated Dirac equations with α-weight are studied. First, decomposition theorems of biaxial symmetric solutions to iterated Dirac equations with α-weight are presented. Then, using the decomposition theorems, some relationships between RHBVPs for biaxial symmetric solutions to iterated Dirac equations with α-weight and RHBVPs for biaxial symmetric monogenic functions are established. Finally, solutions to RHBVPs for biaxial symmetric solutions to iterated Dirac equations with α-weight are given.

Stability threshold of nearly-Couette shear flows with Navier boundary conditions in 2D

Jacob Bedrossian, Siming He, Sameer Iyer +1 more

math.AP · · cited 2× · arXiv ↗ · more like this ·

In this work, we prove a threshold theorem for the 2D Navier-Stokes equations posed on the periodic channel, $\mathbb{T} \times [-1,1]$, supplemented with Navier boundary conditions $ω|_{y = \pm 1} = 0$. Initial datum is taken to be a perturbation of Couette in the following sense: the shear component of the perturbation is assumed small (in an appropriate Sobolev space) but importantly is independent of $ν$. On the other hand, the nonzero modes are assumed size $O(ν^{\frac12})$ in an anisotropic Sobolev space. For such datum, we prove nonlinear enhanced dissipation and inviscid damping for the resulting solution. The principal innovation is to capture quantitatively the \textit{inviscid damping}, for which we introduce a new Singular Integral Operator which is a physical space analogue of the usual Fourier multipliers which are used to prove damping. We then include this SIO in the…

On the $8π$-critical mass threshold of a Patlak-Keller-Segel-Navier-Stokes system

Yishu Gong, Siming He

math.AP · · cited 1× · arXiv ↗ · more like this ·

In this paper, we proposed a coupled Patlak-Keller-Segel-Navier-Stokes system, which has dissipative free energy. On the plane $\rr^2$, if the total mass of the cells is strictly less than $8π$, classical solutions exist for any finite time, and their $H^s$-Sobolev norms are almost uniformly bounded in time. For the radially symmetric solutions, this $8π$-mass threshold is critical. On the torus $\mathbb{T}^2$, the solutions are uniformly bounded in time under the same mass constraint.

Nonlinear Inviscid Damping of Couette Flow with Vorticity Perturbations Supported up to the Boundary

Siming He, Sameer Iyer, Weiren Zhao

math.AP · · cited 0× · arXiv ↗ · more like this ·

We establish nonlinear inviscid damping of the near-Couette flows in a finite channel, with vorticity perturbations vanishing to infinite order at the boundary. To study the boundary effect, we decompose the vorticity into dyadic layers whose supports expand toward the boundary, and use an inductive Newton scheme to track nonlinear interactions across layers. Finally, precise elliptic estimates are established to quantify the velocity generated by different layers.

Fast separation of Euler and Navier--Stokes flows in two dimensions

Siming He, Yupei Huang, Jincheng Yang

math.AP · · cited 0× · arXiv ↗ · more like this ·

We study the long-time validity of the inviscid limit problem by measuring when the two-dimensional ideal fluid ceases to accurately approximate a slightly viscous fluid with the same initial datum. We construct a stationary solution of the Euler equation from which the Navier--Stokes solution separates at a time scale proportional to the logarithm of the inverse viscosity, achieving the fastest separation from Euler steady states.

BT-Agent: Deploying Onboard Robot Agents with Behavior Trees and Multi-Fidelity Evaluation

Fabio Hübel, Adam Lilja, Yue Yin +5 more

cs.RO · · cited 0× · arXiv ↗ · more like this ·

Deploying language-model agents onboard mobile robots requires task reasoning and responsive execution within limited compute and memory. We propose BT-Agent, an onboard architecture combining LLM-driven task decisions with predefined behavior-tree (BT) skills. These skills enforce execution conditions and return structured failure feedback. We demonstrate fully onboard operation with a 9B model on a quadruped across five task categories. Intervention tests show that local checks stop ongoing navigation without waiting for another model decision, while the agent remains active to retry or report constraints. To compare agent designs efficiently, we introduce a 161-task multi-fidelity benchmark, evaluated in text on all tasks and in Habitat on 39. Text supports 10.9$\times$ higher evaluation throughput, while simulation reveals how perception and motion failures change task outcomes.…

A Scene Language Model for Open-Vocabulary Scene Mapping

Adam Lilja, Fabio Hübel, Siming He +6 more

cs.CV · · cited 0× · arXiv ↗ · more like this ·

Open-vocabulary 3D scene mapping aims to build a persistent representation of the objects in an environment. Existing systems typically rely on engineered mapping pipelines to associate observations, merge information across views, and maintain a consistent scene representation over time. Many additionally store feature-rich object representations, such as embeddings or image crops, increasing the size and complexity of the persistent memory. We introduce SceneLM, a Scene-Language Model that directly maintains a textual scene map. The full scene is represented as a structured text list of objects, which serves as the model's only persistent memory. For each input image, the model reads the current scene state and updates the map by adding, editing, and removing objects. To learn this behavior, we introduce supervision tasks for iterative scene map maintenance together with an automatic…