arXiv Science⌕ Search

arXiv subjects

Le Chen

Publications and source records attributed to Le Chen.

At least 91 records · Page 5Linked to original sources

Uncertainty Guided Policy for Active Robotic 3D Reconstruction using Neural Radiance Fields

In this paper, we tackle the problem of active robotic 3D reconstruction of an object. In particular, we study how a mobile robot with an arm-held camera can select a favorable number of views to recover an object's 3D shape efficiently. Contrary to the existing solution to this problem, we leverage the popular neural radiance fields-based object representation, which has recently shown impressive results for various computer vision tasks. However, it is not straightforward to directly reason about an object's explicit 3D geometric details using such a representation, making the next-best-view selection problem for dense 3D reconstruction challenging. This paper introduces a ray-based volumetric uncertainty estimator, which computes the entropy of the weight distribution of the color samples along each ray of the object's implicit neural representation. We show that it is possible to infer the uncertainty of the underlying 3D geometry given a novel view with the proposed estimator. We then present a next-best-view selection policy guided by the ray-based volumetric uncertainty in neural radiance fields-based representations. Encouraging experimental results on synthetic and real-world data suggest that the approach presented in this paper can enable a new research direction of using an implicit 3D object representation for the next-best-view problem in robot vision applications, distinguishing our approach from the existing approaches that rely on explicit 3D geometric modeling.

cs.CV↗

Invariant measures for the nonlinear stochastic heat equation with no drift term

This paper deals with the long term behavior of the solution to the nonlinear stochastic heat equation $\partial u /\partial t - \frac{1}{2}Δu = b(u)\dot{W}$, where $b$ is assumed to be a globally Lipschitz continuous function and the noise $\dot{W}$ is a centered and spatially homogeneous Gaussian noise that is white in time. Using the moment formulas obtained in [9, 10], we identify a set of conditions on the initial data, the correlation measure and the weight function $ρ$, which will together guarantee the existence of an invariant measure in the weighted space $L^2_ρ(\mathbb{R}^d)$. In particular, our result includes the parabolic Anderson model (i.e., the case when $b(u) = λu$) starting from the Dirac delta measure.

math.PR↗

Superlinear stochastic heat equation on $\mathbb{R}^d$

In this paper, we study the stochastic heat equation (SHE) on $\mathbb{R}^d$ subject to a centered Gaussian noise that is white in time and colored in space. We establish the existence and uniqueness of the random field solution in the presence of locally Lipschitz drift and diffusion coefficients, which can have certain superlinear growth. This is a nontrivial extension of the recent work by Dalang, Khoshnevisan and Zhang (2019), where the one-dimensional SHE on $[0,1]$ subject to space-time white noise has been studied.

math.PR↗

Parabolic Anderson model with rough noise in space and rough initial conditions

In this note, we consider the parabolic Anderson model on $\mathbb{R}_{+} \times \mathbb{R}$, driven by a Gaussian noise which is fractional in time with index $H_0>1/2$ and fractional in space with index $0 3/4$. Under a general condition on the initial data, we prove the existence and uniqueness of the mild solution and obtain its exponential upper bounds in time for all $p$-th moments with $p\ge 2$.

math.PR↗

Moments and asymptotics for a class of SPDEs with space-time white noise

In this article, we consider the nonlinear stochastic partial differential equation of fractional order in both space and time variables with constant initial condition: \begin{equation*} \left(\partial^β_t+\dfracν{2}\left(-Δ\right)^{α/ 2}\right) u(t, x)= ~ I_{t}^γ\left[λu(t, x) \dot{W}(t, x)\right] \quad t>0,~ x\in\mathbb R^d, \end{equation*} where $\dot{W}$ is space-time white noise, $α>0$, $β\in(0,2]$, $γ\ge 0$, $λ\neq0$ and $ν>0$. The existence and uniqueness of solution in the Itô-Skorohod sense is obtained under Dalang's condition. We obtain explicit formulas for both the second moment and the second moment Lyapunov exponent. We derive the $p$-th moment upper bounds and find the matching lower bounds. Our results solve a large class of conjectures regarding the order of the $p$-th moment Lyapunov exponents. In particular, by letting $β=2$, $α=2$, $γ=0$, and $d=1$, we confirm the following standing conjecture for the stochastic wave equation: \begin{align*} t^{-1}\log\mathbb E[u(t,x)^p] \asymp p^{3/2}, \quad \text{for $p\ge 2$ as $t\to \infty$.} \end{align*} The method for the lower bounds is inspired by a recent work by Hu and Wang [HW21], where the authors focus on the space-time colored Gaussian noise.

math.PR↗

Central limit theorems for parabolic stochastic partial differential equations

Let $\{u(t\,,x)\}_{t\ge 0, x\in \mathbb{R}^d}$ denote the solution of a $d$-dimensional nonlinear stochastic heat equation that is driven by a Gaussian noise, white in time with a homogeneous spatial covariance that is a finite Borel measure $f$ and satisfies Dalang's condition. We prove two general functional central limit theorems for occupation fields of the form $N^{-d} \int_{\mathbb{R}^d} g(u(t\,,x)) ψ(x/N)\, \mathrm{d} x$ as $N\rightarrow \infty$, where $g$ runs over the class of Lipschitz functions on $\mathbb{R}^d$ and $ψ\in L^2(\mathbb{R}^d)$. The proof uses Poincaré-type inequalities, Malliavin calculus, compactness arguments, and Paul Lévy's classical characterization of Brownian motion as the only mean zero, continuous Lévy process. Our result generalizes central limit theorems of Huang et al \cite{HuangNualartViitasaari2018,HuangNualartViitasaariZheng2019} valid when $g(u)=u$ and $ψ= \mathbf{1}_{[0,1]^d}$.

math.PR↗

Unified Data Collection for Visual-Inertial Calibration via Deep Reinforcement Learning

Visual-inertial sensors have a wide range of applications in robotics. However, good performance often requires different sophisticated motion routines to accurately calibrate camera intrinsics and inter-sensor extrinsics. This work presents a novel formulation to learn a motion policy to be executed on a robot arm for automatic data collection for calibrating intrinsics and extrinsics jointly. Our approach models the calibration process compactly using model-free deep reinforcement learning to derive a policy that guides the motions of a robotic arm holding the sensor to efficiently collect measurements that can be used for both camera intrinsic calibration and camera-IMU extrinsic calibration. Given the current pose and collected measurements, the learned policy generates the subsequent transformation that optimizes sensor calibration accuracy. The evaluations in simulation and on a real robotic system show that our learned policy generates favorable motion trajectories and collects enough measurements efficiently that yield the desired intrinsics and extrinsics with short path lengths. In simulation we are able to perform calibrations 10 times faster than hand-crafted policies, which transfers to a real-world speed up of 3 times over a human expert.

cs.RO↗

Interpolating the Stochastic Heat and Wave Equations with Time-independent Noise: Solvability and Exact Asymptotics

In this article, we study a class of stochastic partial differential equations with fractional differential operators subject to some time-independent multiplicative Gaussian noise. We derive sharp conditions, under which a unique global $L^p(Ω)$-solution exits for all $p\ge 2$. In this case, we derive exact moment asymptotics following the same strategy in a recent work by Balan et al [1]. In the case when there exits only a local solution, we determine the precise deterministic time, $T_2$, before which a unique $L^2(Ω)$-solution exits, but after which the series corresponding to the $L^2(Ω)$ moment of the solution blows up. By properly choosing the parameters, results in this paper interpolate the known results for both stochastic heat and wave equations.

math.PR↗

Multi-scale super-resolution generation of low-resolution scanned pathological images

Background. Digital pathology has aroused widespread interest in modern pathology. The key of digitalization is to scan the whole slide image (WSI) at high magnification. The lager the magnification is, the richer details WSI will provide, but the scanning time is longer and the file size of obtained is larger. Methods. We design a strategy to scan slides with low resolution (5X) and a super-resolution method is proposed to restore the image details when in diagnosis. The method is based on a multi-scale generative adversarial network, which sequentially generates three high-resolution images such as 10X, 20X and 40X. Results. The peak-signal-to-noise-ratio of 10X to 40X generated images are 24.16, 22.27 and 20.44, and the structural-similarity-index are 0.845, 0.680 and 0.512, which are better than other super-resolution networks. Visual scoring average and standard deviation from three pathologists is 3.63 plus-minus 0.52, 3.70 plus-minus 0.57 and 3.74 plus-minus 0.56 and the p value of analysis of variance is 0.37, indicating that generated images include sufficient information for diagnosis. The average value of Kappa test is 0.99, meaning the diagnosis of generated images is highly consistent with that of the real images. Conclusion. This proposed method can generate high-quality 10X, 20X, 40X images from 5X images at the same time, in which the time and storage costs of digitalization can be effectively reduced up to 1/64 of the previous costs. The proposed method provides a better alternative for low-cost storage, faster image share of digital pathology. Keywords. Digital pathology; Super-resolution; Low resolution scanning; Low cost

eess.IV↗

Exact asymptotics of the stochastic wave equation with time-independent noise

In this article, we study the stochastic wave equation in all dimensions $d\leq 3$, driven by a Gaussian noise $\dot{W}$ which does not depend on time. We assume that either the noise is white, or the covariance function of the noise satisfies a scaling property similar to the Riesz kernel. The solution is interpreted in the Skorohod sense using Malliavin calculus. We obtain the exact asymptotic behaviour of the $p$-th moment of the solution either when the time is large or when $p$ is large. For the critical case, that is the case when $d=3$ and the noise is white, we obtain the exact transition time for the second moment to be finite.

math.PR↗

Hölder regularity of the nonlinear stochastic time-fractional slow and fast diffusion equations on $\mathbb{R}^d$

In this paper, we use local fraction derivative to show the Hölder continuity of the solution to the following nonlinear time-fractional slow and fast diffusion equation: $$\left(\partial^β+\fracν{2}(-Δ)^{α/2}\right)u(t,x) = I_t^γ\left[σ\left(u(t,x)\right)\dot{W}(t,x)\right],\quad t>0,\: x\in\mathbb{R}^d,$$ where $\dot{W}$ is the space-time white noise, $α\in(0,2]$, $β\in(0,2)$, $γ\ge 0$ and $ν>0$, under the condition that $2(β+γ)-1-dβ/α>0$. The case when $β+γ\le 1$ has been obtained in \cite{ChHuNu19}. In this paper, we have removed this extra condition, which in particular includes all cases for $β\in(0,2)$.

math.PR↗

Learning Trajectories for Visual-Inertial System Calibration via Model-based Heuristic Deep Reinforcement Learning

Visual-inertial systems rely on precise calibrations of both camera intrinsics and inter-sensor extrinsics, which typically require manually performing complex motions in front of a calibration target. In this work we present a novel approach to obtain favorable trajectories for visual-inertial system calibration, using model-based deep reinforcement learning. Our key contribution is to model the calibration process as a Markov decision process and then use model-based deep reinforcement learning with particle swarm optimization to establish a sequence of calibration trajectories to be performed by a robot arm. Our experiments show that while maintaining similar or shorter path lengths, the trajectories generated by our learned policy result in lower calibration errors compared to random or handcrafted trajectories.

cs.RO↗

Central limit theorems for spatial averages of the stochastic heat equation via Malliavin-Stein's method

Suppose that $\{u(t\,, x)\}_{t >0, x \in\mathbb{R}^d}$ is the solution to a $d$-dimensional stochastic heat equation driven by a Gaussian noise that is white in time and has a spatially homogeneous covariance that satisfies Dalang's condition. The purpose of this paper is to establish quantitative central limit theorems for spatial averages of the form $N^{-d} \int_{[0,N]^d} g(u(t\,,x))\, \mathrm{d} x$, as $N\rightarrow\infty$, where $g$ is a Lipschitz-continuous function or belongs to a class of locally-Lipschitz functions, using a combination of the Malliavin calculus and Stein's method for normal approximations. Our results include a central limit theorem for the {\it Hopf-Cole} solution to KPZ equation. We also establish a functional central limit theorem for these spatial averages.

math.PR↗

Stochastic comparisons for stochastic heat equation

We establish the stochastic comparison principles, including moment comparison principle as a special case, for solutions to the following nonlinear stochastic heat equation on $\mathbb{R}^d$ \[ \left(\frac{\partial }{\partial t} -\frac{1}{2}Δ\right) u(t,x) = ρ(u(t,x)) \:\dot{M}(t,x), \] where $\dot{M}$ is a spatially homogeneous Gaussian noise that is white in time and colored in space, and $ρ$ is a Lipschitz continuous function that vanishes at zero. These results are obtained for rough initial data and under Dalang's condition, namely, $\int_{\mathbb{R}^d}(1+|ξ|^2)^{-1}\hat{f}(\text{d} ξ)<\infty$, where $\hat{f}$ is the spectral measure of the noise. We establish the comparison principles by comparing either the diffusion coefficient $ρ$ or the correlation function of the noise $f$. As corollaries, we obtain Slepian's inequality for SPDEs and SDEs.

math.PR↗

Spatial ergodicity for SPDEs via Poincaré-type inequalities

Consider a parabolic stochastic PDE of the form $\partial_t u=\frac{1}{2}Δu + σ(u)η$, where $u=u(t\,,x)$ for $t\ge0$ and $x\in\mathbb{R}^d$, $σ:\mathbb{R}\rightarrow\mathbb{R}$ is Lipschitz continuous and non random, and $η$ is a centered Gaussian noise that is white in time and colored in space, with a possibly-signed homogeneous spatial correlation $f$. If, in addition, $u(0)\equiv1$, then we prove that, under a mild decay condition on $f$, the process $x\mapsto u(t\,,x)$ is stationary and ergodic at all times $t>0$. It has been argued that, when coupled with moment estimates, spatial ergodicity of $u$ teaches us about the intermittent nature of the solution to such SPDEs \cite{BertiniCancrini1995,KhCBMS}. Our results provide rigorous justification of such discussions. Our methods hinge on novel facts from harmonic analysis and functions of positive type, as well as from Malliavin calculus and Poincaré inequalities. We further showcase the utility of these Poincaré inequalities by: (a) describing conditions that ensure that the random field $u(t)$ is mixing for every $t>0$; and by (b) giving a quick proof of a conjecture of Conus et al \cite{CJK12} about the "size" of the intermittency islands of $u$. The ergodicity and the mixing results of this paper are sharp, as they include the classical theory of Maruyama \cite{Maruyama} (see also Dym and McKean \cite{DymMcKean}) in the simple setting where the nonlinear term $σ$ is a constant function.

math.PR↗

Spatial ergodicity for SPDEs via a Poincaré-type inequality

Consider a parabolic stochastic PDE of the form $\partial_t u=\frac{1}{2}Δu + σ(u)η$, where $u=u(t\,,x)$ for $t\ge0$ and $x\in\mathbb{R}^d$, $σ:\mathbb{R}\to\mathbb{R}$ is Lipschitz continuous and non random, and $η$ is a centered Gaussian noise that is white in time and colored in space, with a possibly-signed homogeneous spatial correlation function $f$. If, in addition, $u(0)\equiv1$, then we prove that, under a mild decay condition on $f$, the process $x\mapsto u(t\,,x)$ is stationary and ergodic at all times $t>0$. It has been argued that, when coupled with moment estimates, spatial ergodicity of $u$ teaches us about the intermittent nature of the solution to such SPDEs \cite{BertiniCancrini1995,KhCBMS}. Our results provide rigorous justification of of such discussions. The proof rests on novel facts about functions of positive type, and on strong localization bounds for comparison of SPDEs.

math.PR↗

Regularity and strict positivity of densities for the stochastic heat equation on $\mathbb{R}^d$

In this paper, we study the stochastic heat equation with a general multiplicative Gaussian noise that is white in time and colored in space. Both regularity and strict positivity of the densities of the solution have been established. The difficulty, and hence the contribution, of the paper lie in three aspects, which include rough initial conditions, degenerate diffusion coefficient, and weakest possible assumptions on the correlation function of the noise. In particular, our results cover the parabolic Anderson model starting from a Dirac delta initial measure.

math.PR↗