arXiv subjects

Jie Xiang

Publications and source records attributed to Jie Xiang.

At least 19 recordsLinked to original sources

arXiv2026

Averaging principles for nonautonomous multiscale McKean-Vlasov stochastic systems

This paper investigates a class of nonautonomous multiscale McKean-Vlasov stochastic systems. By leveraging the nonautonomous Poisson equation, we rigorously establish both strong and weak averaging principles, accompanied by explicit convergence rates. Notably, the coefficients of the averaging equations derived in the general case retain dependence on the scaling parameter $\varepsilon$. However, under the additional assumptions that the fast-scale coefficients are either asymptotically convergent or time-periodic, we demonstrate that the slow component converges, in the strong or weak sense, to averaging equations with coefficients independent of $\varepsilon$.

math.PR↗

arXiv2026

Ascend-RaBitQ: Heterogeneous NPU-CPU Acceleration of Billion-Scale Similarity Search with 1-bit Quantization

Vector similarity search is a critical component of modern AI systems, but traditional CPU-based implementations face fundamental scalability bottlenecks for billion-scale corpora due to prohibitive computational overhead and memory bandwidth limitations. While Neural Processing Units (NPUs) offer orders-of-magnitude higher compute density, existing CPU/GPU-optimized 1-bit RaBitQ quantization implementations cannot be directly ported to NPU architectures due to fundamental hardware mismatches, and homogeneous design paradigms struggle to simultaneously balance accuracy, memory footprint, and performance. This paper presents Ascend-RaBitQ, the first heterogeneous NPU-CPU optimized IVF-RaBitQ system for billion-scale vector search, built on the core insight that decoupling coarse ranking (NPU) from fine ranking (CPU) allows each stage to leverage its optimal hardware, breaking the long-standing accuracy-memory-performance trade-off. We propose a three-stage heterogeneous execution path comprising AI Core-accelerated coarse ranking on 1-bit quantized vectors, on-device AI CPU Top-k processing, and host CPU fine re-ranking on full-precision vectors. We introduce four NPU architecture-native optimizations: fused AIC-AIV operators for parallel distance computation, computation flow restructuring to exploit rotation orthogonality, fine-grained index block-level load balancing that breaks query boundaries, and intra-NPU pipeline parallelism between AI Core and AI CPU to mask Top-k latency. Evaluation on standard datasets shows that Ascend-RaBitQ achieves 3.0X to 62.8X faster index construction than the CPU baseline, up to 11.7X throughput improvement over the fastest CPU IVF-RaBitQ implementation, and over two orders of magnitude over the mathematically equivalent CPU baseline, while demonstrating encouraging scalability on distributed multi-NPU systems.

cs.IR↗

arXiv2025

DriveLiDAR4D: Sequential and Controllable LiDAR Scene Generation for Autonomous Driving

The generation of realistic LiDAR point clouds plays a crucial role in the development and evaluation of autonomous driving systems. Although recent methods for 3D LiDAR point cloud generation have shown significant improvements, they still face notable limitations, including the lack of sequential generation capabilities and the inability to produce accurately positioned foreground objects and realistic backgrounds. These shortcomings hinder their practical applicability. In this paper, we introduce DriveLiDAR4D, a novel LiDAR generation pipeline consisting of multimodal conditions and a novel sequential noise prediction model LiDAR4DNet, capable of producing temporally consistent LiDAR scenes with highly controllable foreground objects and realistic backgrounds. To the best of our knowledge, this is the first work to address the sequential generation of LiDAR scenes with full scene manipulation capability in an end-to-end manner. We evaluated DriveLiDAR4D on the nuScenes and KITTI datasets, where we achieved an FRD score of 743.13 and an FVD score of 16.96 on the nuScenes dataset, surpassing the current state-of-the-art (SOTA) method, UniScene, with an performance boost of 37.2% in FRD and 24.1% in FVD, respectively.

cs.CV↗

arXiv2025

Asymptotic behaviors of multiscale McKean-Vlasov stochastic systems

In this paper, we investigate a class of multiscale McKean-Vlasov stochastic systems, where the entire system depends on the distributions of both fast and slow components. First of all, by applying the Poisson equation method, we prove that the slow component converges to the solution of the averaging equation in the $L^p$ ($p\geq 2$) space with the optimal convergence order $\frac12$. Then we establish a central limit theorem for these systems and derive the weak convergence rate using the Poisson equation technique and the regularity properties of the associated Cauchy problem.

math.PR↗

arXiv2025

Comorbidity-Informed Transfer Learning for Neuro-developmental Disorder Diagnosis

Neuro-developmental disorders are manifested as dysfunctions in cognition, communication, behaviour and adaptability, and deep learning-based computer-aided diagnosis (CAD) can alleviate the increasingly strained healthcare resources on neuroimaging. However, neuroimaging such as fMRI contains complex spatio-temporal features, which makes the corresponding representations susceptible to a variety of distractions, thus leading to less effective in CAD. For the first time, we present a Comorbidity-Informed Transfer Learning(CITL) framework for diagnosing neuro-developmental disorders using fMRI. In CITL, a new reinforced representation generation network is proposed, which first combines transfer learning with pseudo-labelling to remove interfering patterns from the temporal domain of fMRI and generates new representations using encoder-decoder architecture. The new representations are then trained in an architecturally simple classification network to obtain CAD model. In particular, the framework fully considers the comorbidity mechanisms of neuro-developmental disorders and effectively integrates them with semi-supervised learning and transfer learning, providing new perspectives on interdisciplinary. Experimental results demonstrate that CITL achieves competitive accuracies of 76.32% and 73.15% for detecting autism spectrum disorder and attention deficit hyperactivity disorder, respectively, which outperforms existing related transfer learning work for 7.2% and 0.5% respectively.

cs.LG↗

arXiv2025

A Confounding Factors-Inhibition Adversarial Learning Framework for Multi-site fMRI Mental Disorder Identification

In open data sets of functional magnetic resonance imaging (fMRI), the heterogeneity of the data is typically attributed to a combination of factors, including differences in scanning procedures, the presence of confounding effects, and population diversities between multiple sites. These factors contribute to the diminished effectiveness of representation learning, which in turn affects the overall efficacy of subsequent classification procedures. To address these limitations, we propose a novel multi-site adversarial learning network (MSalNET) for fMRI-based mental disorder detection. Firstly, a representation learning module is introduced with a node information assembly (NIA) mechanism to better extract features from functional connectivity (FC). This mechanism aggregates edge information from both horizontal and vertical directions, effectively assembling node information. Secondly, to generalize the feature across sites, we proposed a site-level feature extraction module that can learn from individual FC data, which circumvents additional prior information. Lastly, an adversarial learning network is proposed as a means of balancing the trade-off between individual classification and site regression tasks, with the introduction of a novel loss function. The proposed method was evaluated on two multi-site fMRI datasets, i.e., Autism Brain Imaging Data Exchange (ABIDE) and ADHD-200. The results indicate that the proposed method achieves a better performance than other related algorithms with the accuracy of 75.56 and 68.92 in ABIDE and ADHD-200 datasets, respectively. Furthermore, the result of the site regression indicates that the proposed method reduces site variability from a data-driven perspective. The most discriminative brain regions revealed by NIA are consistent with statistical findings, uncovering the "black box" of deep learning to a certain extent.

cs.LG↗

arXiv2025

Averaging principles and central limit theorems for multiscale McKean-Vlasov stochastic systems

In this paper, we study a class of multiscale McKean-Vlasov stochastic systems where the entire system depends on the distribution of the fast component. First of all, by the Poisson equation method we prove that the slow component converges to the solution of the averaging equation in the $L^p$ ($p\geq 2$) space with the optimal convergence rate 1/2. Then a central limit theorem is established by tightness.

math.PR↗

arXiv2024

DiVE: DiT-based Video Generation with Enhanced Control

Generating high-fidelity, temporally consistent videos in autonomous driving scenarios faces a significant challenge, e.g. problematic maneuvers in corner cases. Despite recent video generation works are proposed to tackcle the mentioned problem, i.e. models built on top of Diffusion Transformers (DiT), works are still missing which are targeted on exploring the potential for multi-view videos generation scenarios. Noticeably, we propose the first DiT-based framework specifically designed for generating temporally and multi-view consistent videos which precisely match the given bird's-eye view layouts control. Specifically, the proposed framework leverages a parameter-free spatial view-inflated attention mechanism to guarantee the cross-view consistency, where joint cross-attention modules and ControlNet-Transformer are integrated to further improve the precision of control. To demonstrate our advantages, we extensively investigate the qualitative comparisons on nuScenes dataset, particularly in some most challenging corner cases. In summary, the effectiveness of our proposed method in producing long, controllable, and highly consistent videos under difficult conditions is proven to be effective.

cs.CV↗

arXiv2023

Exploring the Mutual Influence between Self-Supervised Single-Frame and Multi-Frame Depth Estimation

Although both self-supervised single-frame and multi-frame depth estimation methods only require unlabeled monocular videos for training, the information they leverage varies because single-frame methods mainly rely on appearance-based features while multi-frame methods focus on geometric cues. Considering the complementary information of single-frame and multi-frame methods, some works attempt to leverage single-frame depth to improve multi-frame depth. However, these methods can neither exploit the difference between single-frame depth and multi-frame depth to improve multi-frame depth nor leverage multi-frame depth to optimize single-frame depth models. To fully utilize the mutual influence between single-frame and multi-frame methods, we propose a novel self-supervised training framework. Specifically, we first introduce a pixel-wise adaptive depth sampling module guided by single-frame depth to train the multi-frame model. Then, we leverage the minimum reprojection based distillation loss to transfer the knowledge from the multi-frame depth network to the single-frame network to improve single-frame depth. Finally, we regard the improved single-frame depth as a prior to further boost the performance of multi-frame depth estimation. Experimental results on the KITTI and Cityscapes datasets show that our method outperforms existing approaches in the self-supervised monocular setting.

cs.CV↗

arXiv2023

Stochastic averaging principle and stability for multi-valued McKean-Vlasov stochastic differential equations with jumps

In this paper, we consider the stochastic averaging principle and stability for multi-valued McKean-Vlasov stochastic differential equations with jumps. First, under certain averaging conditions, we are able to show that the solutions of the equations concerned can be approximated by solutions of the associated averaged multi-valued McKean-Vlasov stochastic differential equations with jumps in the sense of the mean square convergence. Second, we extend the classical Itô's formula from stochastic differential equations to multi-valued McKean-Vlasov stochastic differential equations with jumps. Last, as application of Itô's formula, we present the exponential stability of second moments, the exponentially 2-ultimate boundedness and the almost surely asymptotic stability for their solutions in terms of a Lyapunov function.

math.PR↗

arXiv2023

Multiple skew orthogonal polynomials and 2-component Pfaff lattice hierarchy

In this paper, we introduce multiple skew-orthogonal polynomials and investigate their connections with classical integrable systems. By using Pfaffian techniques, we show that multiple skew-orthogonal polynomials can be expressed by multi-component Pfaffian tau-functions upon appropriate deformations. Moreover, a two-component Pfaff lattice hierarchy, which is equivalent to the Pfaff-Toda hierarchy studied by Takasaki, is obtained by considering the recurrence relations and Cauchy transforms of multiple skew-orthogonal polynomials.

math-ph↗

arXiv2022

Visual Attention-based Self-supervised Absolute Depth Estimation using Geometric Priors in Autonomous Driving

Although existing monocular depth estimation methods have made great progress, predicting an accurate absolute depth map from a single image is still challenging due to the limited modeling capacity of networks and the scale ambiguity issue. In this paper, we introduce a fully Visual Attention-based Depth (VADepth) network, where spatial attention and channel attention are applied to all stages. By continuously extracting the dependencies of features along the spatial and channel dimensions over a long distance, VADepth network can effectively preserve important details and suppress interfering features to better perceive the scene structure for more accurate depth estimates. In addition, we utilize geometric priors to form scale constraints for scale-aware model training. Specifically, we construct a novel scale-aware loss using the distance between the camera and a plane fitted by the ground points corresponding to the pixels of the rectangular area in the bottom middle of the image. Experimental results on the KITTI dataset show that this architecture achieves the state-of-the-art performance and our method can directly output absolute depth without post-processing. Moreover, our experiments on the SeasonDepth dataset also demonstrate the robustness of our model to multiple unseen environments.

cs.CV↗

arXiv2020

Machine Learning based Pallets Detection and Tracking in AGVs

The use of automated guided vehicles (AGVs) has played a pivotal role in manufacturing and distribution operations, providing reliable and efficient product handling. In this project, we constructed a deep learning-based pallets detection and tracking architecture for pallets detection and position tracking. By using data preprocessing and augmentation techniques and experiment with hyperparameter tuning, we achieved the result with 25% reduction of error rate, 28.5% reduction of false negative rate, and 20% reduction of training time.

cs.CV↗

arXiv2017

Corner singularity and its application in regular parameters optimization: technique renovation for Grad-Shafranov reconstruction

Further studies on the corner singularity of GS reconstruction are are compiled in this paper. It's focused on solution of the Data Completion (DC) problem with the Extended Hilbert Transform (EHT) over plane rectangular region. Optimal selections of the regular parameters in{\it Tikhonov} solution of corresponding DC problem are developed in this study. The 6-parameter regular solutions and the {\it Jacobian} matrix and one {\it Hessian} tensor to the regular parameters are derived in this work. A concise formula for EHT in the near field of corners, which shows property for EHT near the corner, are also provided. It serves as the additional constraints for our parameter optimization problem (OP). Third, a nonlinear convex function defined by the regular solution and the corner constraints is introduced and is taken as the object function for the OP of the 6 regular parameters on half-space $(\mathbf{p}>0)$. Given an initial guess of $\mathbf{p}_0$, the optimal parameters are solved from the OP through a well known constrained nonlinear optimization method. Last, the benchmark tests to the proposed solution approach are carried out, and detailed results from totally 9 different bench-cases are tabulated. In contrast to solutions with given regular parameters, our bench results demonstrate that an objective way for selection of the optimal $\mathbf{p}$ is successfully laid out here. Robustness and efficiency of the suggested new approach are also highlighted in this study.

astro-ph.IM↗

arXiv2016

Comparative analysis of anisotropic material properties of uniaxial nematics formed by flexible dimers and rod-like monomers

We report temperature dependencies of material properties such as dielectric anisotropy, birefringence, splay (K11), twist (K22), and bend (K33) elastic constants of the uniaxial nematic (N) phase formed by flexible dimers of DTC5C9 and compare their behavior to that of a corresponding monomer MCT5. DTC5C9 forms a twist-bend nematic (Ntb) at temperatures below the N phase. Anisotropic properties of MCT5 are typical of the rod-like mesogens. In particular, birefringence increases as the temperature is reduced, following the classic behavior, described by Haller. The elastic constants also follow the standard behavior, with their ratios being practically temperature-independent. In contrast, DTC5C9 shows a dramatic departure from the standard case. Birefringence changes non-monotonously with temperature, decreasing on approaching the N-Ntb phase transition. decreases strongly to 0.4 pN near the N - Ntb transition, although remains finite. The ratios of the elastic constants in DTC5C9 show a strong temperature dependence that can be associated with the bend-induced changes in the orientational distribution function. The measured elastic properties are consistent with the tendency of the dimeric molecules to adopt bent configurations that give rise to the Ntb phase.

cond-mat.soft↗

arXiv2015

Electrically tunable selective reflection of light from ultraviolet to visible and infrared by heliconical cholesterics

Cholesteric liquid crystals with helicoidal molecular architecture are known for their ability to selectively reflect light with the wavelength that is determined by the periodicity of molecular orientations. Here we demonstrate that by using a cholesteric with oblique helicoidal(heliconical) structure, as opposed to the classic right-angle helicoid, one can vary the wavelength of selectively reflected light in a broad spectral range, from ultraviolet to visible and infrared (360-1520 nm for the same chemical composition) by simply adjusting the electric field applied parallel to the helicoidal axis. The effect exists in a wide temperature range (including the room temperatures) and thus can enable many applications that require dynamically controlled transmission and reflection of electromagnetic waves, from energy-saving smart windows to tunable organic lasers, reflective color display, and transparent see-through displays.

cond-mat.soft↗

arXiv2015

Domain Walls and Anchoring Transitions Mimicking Nematic Biaxiality in the Oxadiazole Bent-Core Liquid Crystal C7

We investigate the origin of secondary disclinations that were recently described as a new evidence of a biaxial nematic phase in an oxadiazole bent-core thermotropic liquid crystal C7. With an assortment of optical techniques such as polarizing optical microscopy, LC PolScope, and fluorescence confocal polarizing microscopy, we demonstrate that the secondary disclinations represent non-singular domain walls formed in an uniaxial nematic during the surface anchoring transition, in which surface orientation of the director changes from tangential (parallel to the bounding plates) to tilted. Each domain wall separates two regions with the director tilted in opposite azimuthal directions. At the centre of the wall, the director remains parallel to the bonding plates. The domain walls can be easily removed by applying a modest electric field. The anchoring transition is explained by the balance of (a) the intrinsic perpendicular surface anchoring produced by the polyimide aligning layer and (b) tangential alignment caused by ionic impurities forming electric double layers. The model is supported by the fact that the temperature of the tangential-tilted anchoring transition decreases as the cell thickness increases and as the concentration of ionic species (added salt) increases. We also demonstrate that the surface alignment is strongly affected by thermal degradation of the samples. The study shows that C7 exhibits only a uniaxial nematic phase and demonstrate yet another mechanism (formation of secondary disclinations) by which a uniaxial nematic can mimic a biaxial nematic behaviour.

cond-mat.soft↗

arXiv2014

Liquid crystal-enabled electroosmosis through spatial charge separation in distorted regions as a novel mechanism of electrokinetics

Electrically-controlled dynamics of fluids and particles at microscales is a fascinating area of research with applications ranging from microfluidics and sensing to sorting of biomolecules. The driving mechanisms are electric forces acting on spatially separated charges in an isotropic medium such as water. Here we demonstrate that anisotropic conductivity of liquid crystals enables new mechanism of highly efficient electro-osmosis rooted in space charging of regions with distorted orientation. The electric field acts on these distortion-separated charges to induce liquid crystal-enabled electro-osmosis (LCEO). LCEO velocities grow with the square of the field, which allows one to use an AC field to drive steady flows and to avoid electrode damage. Ionic currents in liquid crystals that have been traditionally considered as an undesirable feature in displays, offer a broad platform for versatile applications such as liquid crystal enabled electrokinetics, micropumping and mixing.

cond-mat.soft↗