arXiv subjects

Xilin Zhou

Publications and source records attributed to Xilin Zhou.

13 recordsLinked to original sources

arXiv2025

Coupling of forward-backward stochastic differential equations on the Wiener space, and application on regularity

S. Geiss and J. Ylinen proposed the coupling method \cite{Geiss:Ylinen:21} to investigate the regularity for the solution to the backward stochastic differential equations with random coefficients. In this paper, we explore this method in setting for the forward-backward stochastic differential equation with random and Lipschitz coefficients, We obtain the regularity in time, and the Malliavin Sobolev ${\mathbb D}_{1,2}$ differentiability for the solution.

math.PR↗

arXiv2025

Homeomorphic Sobolev extensions and integrability of hyperbolic metric

Very recently, it was proved that if the hyperbolic metric of a planar Jordan domain is $L^q$-integrable for some $q\in (1,\infty)$, then every homeomorphic parametrization of the boundary Jordan curve via the unit circle can be extended to a Sobolev homeomorphism of the entire disk. This naturally raises the question of whether the extension holds under more general integrability conditions on the hyperbolic metric. In this work, we examine the case where the hyperbolic metric is $ϕ$-integrable. Under appropriate conditions on the function $ϕ$, we establish the existence of a Sobolev homeomorphic extension for every homeomorphic parametrization of the Jordan curve. Moreover, we demonstrate the sharpness of our result by providing an explicit counterexample.

math.CV↗

arXiv2025

Regularity of stochastic differential equations on the Wiener space by coupling

Using the coupling method introduced in \cite{Geiss:Ylinen:21}, we investigate regularity properties of stochastic differential equations, where we consider the Lipschitz case in $\R^d$ and allow for Hölder continuity of the diffusion coefficient of scalar valued stochastic differential equations. Two cases of the coupling method are of special interest: The uniform coupling to treat the Malliavin Sobolev space $\D_{1,2}$ and real interpolation spaces, and secondly a cut-off coupling to treat the $L_p$-variation of backward stochastic differential equations where the forward process is the investigated stochastic differential equation.

math.PR↗

arXiv2024

Performance assessment of ADAS in a representative subset of critical traffic situations

As a variety of automated collision prevention systems gain presence within personal vehicles, rating and differentiating the automated safety performance of car models has become increasingly important for consumers, manufacturers, and insurers. In 2023, Swiss Re and partners initiated an eight-month long vehicle testing campaign conducted on a recognized UNECE type approval authority and Euro NCAP accredited proving ground in Germany. The campaign exposed twelve mass-produced vehicle models and one prototype vehicle fitted with collision prevention systems to a selection of safety-critical traffic scenarios representative of United States and European Union accident landscape. In this paper, we compare and evaluate the relative safety performance of these thirteen collision prevention systems (hardware and software stack) as demonstrated by this testing campaign. We first introduce a new scoring system which represents a test system's predicted impact on overall real-world collision frequency and reduction of collision impact energy, weighted based on the real-world relevance of the test scenario. Next, we introduce a novel metric that quantifies the realism of the protocol and confirm that our test protocol is a plausible representation of real-world driving. Finally, we find that the prototype system in its pre-release state outperforms the mass-produced (post-consumer-release) vehicles in the majority of the tested scenarios on the test track.

cs.RO↗

arXiv2024

Realistic Extreme Behavior Generation for Improved AV Testing

This work introduces a framework to diagnose the strengths and shortcomings of Autonomous Vehicle (AV) collision avoidance technology with synthetic yet realistic potential collision scenarios adapted from real-world, collision-free data. Our framework generates counterfactual collisions with diverse crash properties, e.g., crash angle and velocity, between an adversary and a target vehicle by adding perturbations to the adversary's predicted trajectory from a learned AV behavior model. Our main contribution is to ground these adversarial perturbations in realistic behavior as defined through the lens of data-alignment in the behavior model's parameter space. Then, we cluster these synthetic counterfactuals to identify plausible and representative collision scenarios to form the basis of a test suite for downstream AV system evaluation. We demonstrate our framework using two state-of-the-art behavior prediction models as sources of realistic adversarial perturbations, and show that our scenario clustering evokes interpretable failure modes from a baseline AV policy under evaluation.

math.OC↗

arXiv2024

Homeomorphic Sobolev extensions of parametrizations of Jordan curves

Each homeomorphic parametrization of a Jordan curve via the unit circle extends to a homeomorphism of the entire plane. It is a natural question to ask if such a homeomorphism can be chosen so as to have some Sobolev regularity. This prompts the simplified question: for a homeomorphic embedding of the unit circle into the plane, when can we find a homeomorphism from the unit disk that has the same boundary values and integrable first-order distributional derivatives? We give the optimal geometric criterion for the interior Jordan domain so that there exists a Sobolev homeomorphic extension for any homeomorphic parametrization of the Jordan curve. The problem is partially motivated by trying to understand which boundary values can correspond to deformations of finite energy.

math.CV↗

arXiv2023

Comparative Safety Performance of Autonomous- and Human Drivers: A Real-World Case Study of the Waymo One Service

This study compares the safety of autonomous- and human drivers. It finds that the Waymo One autonomous service is significantly safer towards other road users than human drivers are, as measured via collision causation. The result is determined by comparing Waymo's third party liability insurance claims data with mileage- and zip-code-calibrated Swiss Re (human driver) private passenger vehicle baselines. A liability claim is a request for compensation when someone is responsible for damage to property or injury to another person, typically following a collision. Liability claims reporting and their development is designed using insurance industry best practices to assess crash causation contribution and predict future crash contributions. In over 3.8 million miles driven without a human being behind the steering wheel in rider-only (RO) mode, the Waymo Driver incurred zero bodily injury claims in comparison with the human driver baseline of 1.11 claims per million miles (cpmm). The Waymo Driver also significantly reduced property damage claims to 0.78 cpmm in comparison with the human driver baseline of 3.26 cpmm. Similarly, in a more statistically robust dataset of over 35 million miles during autonomous testing operations (TO), the Waymo Driver, together with a human autonomous specialist behind the steering wheel monitoring the automation, also significantly reduced both bodily injury and property damage cpmm compared to the human driver baselines.

cs.RO↗

arXiv2022

Low Thermal Conductivity Phase Change Memory Superlattices

Phase change memory devices are typically reset by melt-quenching a material to radically lower its electrical conductance. The high power and concomitantly high current density required to reset phase change materials is the major issue that limits the access times of 3D phase change memory architectures. Phase change superlattices were developed to lower the reset energy by confining the phase transition to the interface between two different phase change materials. However, the high thermal conductivity of the superlattices means that heat is poorly confined within the phase change material, and most of the thermal energy is wasted to the surrounding materials. Here, we identified Ti as a useful dopant for substantially lowering the thermal conductivity of Sb2Te3-GeTe superlattices whilst also stabilising the layered structure from unwanted disordering. We demonstrate via laser heating that lowering the thermal conductivity by doping the Sb2Te3 layers with Ti halves the switching energy compared to superlattices that only use interfacial phase change transitions and strain engineering. The thermally optimized superlattice has (0 0 l) crystallographic orientation yet a thermal conductivity of just 0.25 W/m.K in the "on" (set) state. Prototype phase change memory devices that incorporate this Ti-doped superlattice switch faster and and at a substantially lower voltage than the undoped superlattice. During switching the Ti-doped Sb2Te3 layers remain stable within the superlattice and only the Ge atoms are active and undergo interfacial phase transitions. In conclusion, we show the potential of thermally optimised Sb2Te3-GeTe superlattices for a new generation of energy-efficient electrical and optical phase change memory.

cond-mat.mtrl-sci↗

arXiv2021

Atomic scale control of spin current transmission at interfaces

Spin transmission at ferromagnet/heavy metal interfaces is of vital importance for many spintronic devices. Usually the spin current transmission is limited by the spin mixing conductance and loss mechanisms such as spin memory loss. In order to understand these effects, we study the interface transmission when an insulating interlayer is inserted between the ferromagnet and the heavy metal. For this we measure the inverse spin Hall voltage generated from optically injected spin current pulses as well as the magnitude of the spin pumping using ferromagnetic resonance. From our results we conclude that significant spin memory loss only occurs for 5d metals with less than half filled d-shell.

cond-mat.mes-hall↗

arXiv2020

Crystallization Mechanism Tuned Phase-Change Materials: Quantum Effect on Te-Terminated Boundary

While phase-change materials (PCMs) composed of chalcogenide have different crystallization mechanisms (CM), such as nucleation-dominated Ge2Sb2Te5 (GST) and growth-dominated GeTe (GT), revealing the essential reason of CM as well as the tuned properties is still a long-standing issue. Here, we remarkably find the distinct stability of Te-terminated (111) boundaries (TTB) in different systems, which provides a path to understand the difference in CM. It stems from the quantum effect of molecular orbital theory: the optimal local chemical composition results in the formation of TTB without dangling bonds (DB) in GST but with DB in GT, where DB destabilizes boundary due to its distorted local environment mismatching Oh symmetry of p orbitals. Moreover, the inner vacancy concentration in GST is alterable and controlled by TTB, manifested by the absence of cubic-to-hexagonal transition in carbon-doped GST of small grains and minimized inner vacancy. Finally, the charge transport property (CTP) is controlled by boundary via changing the density of charge or hole nearby as well as vacancy. These findings open the door to tune CTP by CM, which is necessary for achieving low-power and ultrafast devices.

cond-mat.mtrl-sci↗

arXiv2020

Giant transition-state enhancement of quasiparticle spin-Hall effect in an exchange-spin-split superconductor detected by non-local magnon spin-transport

Although recent experiments and theories have shown a variety of exotic transport properties of non-equilibrium quasiparticles (QPs) in superconductor (SC)-based devices with either Zeeman or exchange spin-splitting, how QP interplays with magnon spin currents remains elusive. Here, using non-local magnon spin-transport devices where a singlet SC (Nb) on top of a ferrimagnetic insulator (Y3Fe5O12) serves as a magnon spin detector, we demonstrate that the conversion efficiency of magnon spin to QP charge via inverse spin-Hall effect (iSHE) in such an exchange-spin-split SC can be greatly enhanced by up to 3 orders of magnitude compared with that in the normal state, particularly when its interface superconducting gap matches the magnon spin accumulation. Through systematic measurements with varying the current density and SC thickness, we identify that superconducting coherence peaks and exchange spin-splitting of the QP density-of-states, yielding a larger spin excitation while retaining a modest QP charge-imbalance relaxation, are responsible for the giant QP iSHE. The latter exchange-field-modified QP relaxation is experimentally proved by spatially resolved measurements with varying the separation of electrical contacts on the spin-split Nb.

cond-mat.mes-hall↗

arXiv2018

Wide band gap phase change material tuned visible photonics

Light strongly interacts with structures that are of a similar scale to its wavelength; typically nanoscale features for light in the visible spectrum. However, the optical response of these nanostructures is usually fixed during the fabrication. Phase change materials offer a way to tune the properties of these structures in nanoseconds. Until now, phase change active photonics use materials that strongly absorb visible light, which limits their application in the visible spectrum. In contrast, Stibnite (Sb2S3) is an under-explored phase change material with a band gap that can be tuned in the visible spectrum from 2.0 to 1.7 eV. We deliberately couple this tuneable band gap to an optical resonator such that it responds dramatically in the visible spectrum to Sb2S3 reversible structural phase transitions. We show that this optical response can be triggered both optically and electrically. High speed reprogrammable Sb2S3 based photonic devices, such as those reported here, are likely to have wide applications in future intelligent photonic systems, holographic displays, and micro-spectrometers.

physics.optics↗

arXiv2015

Layered 2D crystals by design: optimisation of Sb$_2$Te$_3$-GeTe van der Waals superlattices

Herein a genetic algorithm for optimising the design of layered 2D heterostructure is proposed. As a proof-of-concept it is applied to Sb$_2$Te$_3$-GeTe phase-change material superlattices, and the resulting lowest energy structure is grown experimentally. The similarity of the computational and experimental structures is verified with the comparison of XRD spectra. The structure is found to be within 0.92 meV/at. from the energetically most favorable known structure for Ge$_2$Sb$_2$Te$_5$.

cond-mat.mtrl-sci↗