Source author record

Guanghan Li

Explore this author’s publications, collaborators and research topics. This catalog record has not been claimed by its author.

ResearcherUnclaimed source record

Catalog footprint

Research overview

11works
5topics
4close collaborators

Published work

11 publications

preprint2024arXiv

Anisotropic flows without global terms and dual Orlicz Christoffel-Minkowski type problem

In this paper, we study the long-time existence and asymptotic behavior for a class of anisotropic non-homogeneous curvature flows without global forcing terms. By the stationary solutions of such anisotropic flows, we obtain existence results for a class of dual Orlicz Christoffel-Minkowski type problems, which is equivalent to solve the PDE G(x,uK,DuK)F(D2uK+uKI)=1G(x,u_K,Du_K)F(D^2u_K+u_KI)=1 on Sn\mathbb S^n for a convex body KK, where DD is the covariant derivative with respect to the standard metric on Sn\mathbb S^n and II is the unit matrix of order nn. This result covers many previous known solutions to LpL^p dual Minkowski problem, LpL^p dual Christoffel-Minkowski problem, and some dual Orlicz Minkowski problem etc.. Meanwhile, the variational formula of some modified quermassintegrals and the corresponding prescribed area measure problem (Orlicz Christoffel-Minkowski type problem) are considered, and inequalities involving modified quermassintegrals are also derived. As corollary, this gives a partial answer about the general prescribed curvature problem raised in Guan-Ren-Wang (CPAM, 2015).

preprint2023arXiv

Cross-Camera Human Motion Transfer by Time Series Analysis

With advances in optical sensor technology, heterogeneous camera systems are increasingly used for high-resolution (HR) video acquisition and analysis. However, motion transfer across multiple cameras poses challenges. To address this, we propose an algorithm based on time series analysis that identifies motion seasonality and constructs an additive model to extract transferable patterns. Validated on real-world data, our algorithm demonstrates effectiveness and interpretability. Notably, it improves pose estimation in low-resolution videos by leveraging patterns derived from HR counterparts, enhancing practical utility. Code is available at: https://github.com/IndigoPurple/TSAMT

preprint2022arXiv

A class of anisotropic inverse Gauss curvature flows and dual Orlicz Minkowski type problem

In this paper, we study the long-time existence and asymptotic behavior for a class of anisotropic inverse Gauss curvature flows. By the stationary solutions of anisotropic flows, we obtain some new existence results for the dual Orlicz Minkowski type problem and even dual Orlicz Minkowski type problem for smooth measures, which is the most reasonable extension of the LpL^p dual Minkowski problem from the dual point of view. The results of corresponding LpL^p versions are LpL^p dual Minkowski problem for p>qp>q; and even LpL^p dual Minkowski problem for p>−1p>-1, or q<1q<1, or some ranges of p<0<qp<0<q, which contain all existence results for smooth measures up to now except p=qp=q or q=n+1q=n+1 (LpL^p Minkowski problem).

preprint2022arXiv

A class of inverse curvature flows and LpL^p dual Christoffel-Minkowski problem

In this paper, we consider a large class of expanding flows of closed, smooth, star-shaped hypersurface in Euclidean space Rn+1\mathbb{R}^{n+1} with speed ψuαρδf−βψu^αρ^δf^{-β}, where ψψ is a smooth positive function on unit sphere, uu is the support function of the hypersurface, ρρ is the radial function, ff is a smooth, symmetric, homogenous of degree one, positive function of the principal curvatures of the hypersurface on a convex cone. When ψ=1ψ=1, we prove that the flow exists for all time and converges to infinity if α+δ+β≤1,β>0α+δ+β\le1, β>0 and α≤0α\le0, while in case α+δ+β>1,α,δ≤0α+δ+β>1,α,δ\le0, the flow blows up in finite time, and where we assume the initial hypersurface to be strictly convex. In both cases the properly rescaled flows converge to a sphere centered the origin. In particular, the results of Gerhardt \cite{GC,GC3} and Urbas \cite{UJ2} can be recovered by putting α=δ=0α=δ=0. Our previous works \cite{DL,DL2} can be recovered by putting δ=0δ=0. By the convergence of these flows, we can give a new proof of uniqueness theorems for solutions to LpL^p-Minkowski problem and LpL^p-Christoffel-Minkowski problem with constant prescribed data. Similarly, we pose the LpL^p dual Christoffel-Minkowski problem and prove a uniqueness theorem for solutions to LpL^p dual Minkowski problem and LpL^p dual Christoffel-Minkowski problem with constant prescribed data. At last, we focus on the longtime existence and convergence of a class of anisotropic flows (i.e. for general function ψψ). The final result not only gives a new proof of many previously known solutions to LpL^p dual Minkowski problem, LpL^p-Christoffel-Minkowski problem, etc. by such anisotropic flows, but also provides solutions to LpL^p dual Christoffel-Minkowski problem with some conditions.

preprint2022arXiv

An efficient nonlinear multigrid solver for the simulation of rarefied gas cavity flow

We study efficient simulation of steady state for rarefied gas flow, which is modeled by the Boltzmann equation with BGK-type collision term. A nonlinear multigrid solver is proposed to resolve the efficiency issue by the following approaches. The unified framework of numerical regularized moment method is first adopted to derive the high-quality discretization of the underlying problem. A fast sweeping iteration is introduced to solve the derived discrete problem more efficiently than the usual time-integration scheme on a single level grid. Taking it as the smoother, the nonlinear multigrid solver is then established to significantly improve the convergence rate. The OpenMP-based parallelization is applied in the implementation to further accelerate the computation. Numerical experiments for two lid-driven cavity flows and a bottom-heated cavity flow are carried out to investigate the performance of the resulting nonlinear multigrid solver. All results show the efficiency and robustness of the solver for both first- and second-order spatial discretization.

preprint2020arXiv

A class of curvature flows expanded by support function and curvature function

In this paper, we consider an expanding flow of closed, smooth, uniformly convex hypersurface in Euclidean \mathbb{R}^{n+1} with speed u^αf^β(α, β\in\mathbb{R}^1), where u is support function of the hypersurface, f is a smooth, symmetric, homogenous of degree one, positive function of the principal curvature radii of the hypersurface. If α\leq 0<β\leq 1-α, we prove that the flow has a unique smooth and uniformly convex solution for all time, and converges smoothly after normalization, to a round sphere centered at the origin.

preprint2020arXiv

Zoom in to the details of human-centric videos

Presenting high-resolution (HR) human appearance is always critical for the human-centric videos. However, current imagery equipment can hardly capture HR details all the time. Existing super-resolution algorithms barely mitigate the problem by only considering universal and low-level priors of im-age patches. In contrast, our algorithm is under bias towards the human body super-resolution by taking advantage of high-level prior defined by HR human appearance. Firstly, a motion analysis module extracts inherent motion pattern from the HR reference video to refine the pose estimation of the low-resolution (LR) sequence. Furthermore, a human body reconstruction module maps the HR texture in the reference frames onto a 3D mesh model. Consequently, the input LR videos get super-resolved HR human sequences are generated conditioned on the original LR videos as well as few HR reference frames. Experiments on an existing dataset and real-world data captured by hybrid cameras show that our approach generates superior visual quality of human body compared with the traditional method.

preprint2013arXiv

Volume-Preserving flow by powers of the mth mean curvature in the hyperbolic space

This paper concerns closed hypersurfaces of dimension n(≥2)n(\geq 2) in the hyperbolic space Hκn+1{\mathbb{H}}_κ^{n+1} of constant sectional curvature κκ evolving in direction of its normal vector, where the speed is given by a power β(≥1/m)β(\geq 1/m) of the mmth mean curvature plus a volume preserving term, including the case of powers of the mean curvature and of the $\mbox{Gauß}$ curvature. The main result is that if the initial hypersurface satisfies that the ratio of the biggest and smallest principal curvature is close enough to 1 everywhere, depending only on nn, mm, ββ and κκ, then under the flow this is maintained, there exists a unique, smooth solution of the flow for all times, and the evolving hypersurfaces exponentially converge to a geodesic sphere of Hκn+1{\mathbb{H}}_κ^{n+1}, enclosing the same volume as the initial hypersurface.

preprint2010arXiv

Mean Curvature Flow of Spacelike Graphs

We prove the mean curvature flow of a spacelike graph in (Σ1×Σ2,g1−g2)(Σ_1\times Σ_2, g_1-g_2) of a map f:Σ1→Σ2f:Σ_1\to Σ_2 from a closed Riemannian manifold (Σ1,g1)(Σ_1,g_1) with Ricci1>0Ricci_1> 0 to a complete Riemannian manifold (Σ2,g2)(Σ_2,g_2) with bounded curvature tensor and derivatives, and with sectional curvatures satisfying K2≤K1K_2\leq K_1, remains a spacelike graph, exists for all time, and converges to a slice at infinity. We also show, with no need of the assumption K2≤K1K_2\leq K_1, that if K1>0K_1>0, or if Ricci1>0Ricci_1>0 and K2≤−cK_2\leq -c, c>0c>0 constant, any map f:Σ1→Σ2f:Σ_1\to Σ_2 is trivially homotopic provided f∗g2<ρg1f^*g_2<ρg_1 where ρ=min⁡Σ1K1/sup⁡Σ2K2+≥0ρ=\min_{Σ_1}K_1/\sup_{Σ_2}K_2^+\geq 0, in case K1>0K_1>0, and ρ=+∞ρ=+\infty in case K2≤0K_2\leq 0. This largely extends some known results for KiK_i constant and Σ2Σ_2 compact, obtained using the Riemannian structure of Σ1×Σ2Σ_1\times Σ_2, and also shows how regularity theory on the mean curvature flow is simpler and more natural in pseudo-Riemannian setting then in the Riemannian one.

preprint2009arXiv

Bernstein-Heinz-Chern results in calibrated manifolds

Given (Mˉ,Ω)(\bar{M},Ω) a calibrated Riemannian manifold with a parallel calibration of rank mm, and MmM^m an immersed orientable submanifold with parallel mean curvature HH we prove that if cos⁡θ\cos θ is bounded away from zero, where θθ is the ΩΩ-angle of MM, and if MM has zero Cheeger constant, then MM is minimal. In the particular case MM is complete with RiccM≥0Ricc^M\geq 0 we may replace the boundedness condition on cos⁡θ\cos θ by cos⁡θ≥Cr−β\cos θ\geq Cr^{-β}, when r→+∞r\to +\infty, where 0≤β<1 0\leqβ<1 and C>0C > 0 are constants and rr is the distance function to a point in MM. Our proof is surprisingly simple and extends to a very large class of submanifolds in calibrated manifolds, in a unified way, the problem started by Heinz and Chern of estimating the mean curvature of graphic hypersurfaces in Euclidean spaces. It is based on a estimation of ∥H∥\|H\| in terms of cos⁡θ\cosθ and an isoperimetric inequality. We also give some conditions to conclude MM is totally geodesic. We study some particular cases.

Research graph

See the researcher in context

Open full explorer

Inspect adjacent papers, topics, institutions and collaborators without losing the researcher page.

Building this map preview

BZPEER is loading the nearby papers, people, topics and institutions for this page.