Searcharxiv⌕ Search

arXiv subjects

Ying Xue

Publications and source records attributed to Ying Xue.

13 recordsLinked to original sources
arXiv2026

DLWM: Dual Latent World Models enable Holistic Gaussian-centric Pre-training in Autonomous Driving

Vision-based autonomous driving has gained much attention due to its low costs and excellent performance. Compared with dense BEV (Bird's Eye View) or sparse query models, Gaussian-centric method is a comprehensive yet sparse representation by describing scene with 3D semantic Gaussians. In this paper, we introduce DLWM, a novel paradigm with Dual Latent World Models specifically designed to enable holistic gaussian-centric pre-training in autonomous driving using two stages. In the first stage, DLWM predicts 3D Gaussians from queries by self-supervised reconstructing multi-view semantic and depth images. Equipped with fine-grained contextual features, in the second stage, two latent world models are trained separately for temporal feature learning, including Gaussian-flow-guided latent prediction for downstream occupancy perception and forecasting tasks, and ego-planning-guided latent prediction for motion planning. Extensive experiments in SurroundOcc and nuScenes benchmarks demonstrate that DLWM shows significant performance gains across Gaussian-centric 3D occupancy perception, 4D occupancy forecasting and motion planning tasks.

cs.CV↗
arXiv2025

Group Inertial Poser: Multi-Person Pose and Global Translation from Sparse Inertial Sensors and Ultra-Wideband Ranging

Tracking human full-body motion using sparse wearable inertial measurement units (IMUs) overcomes the limitations of occlusion and instrumentation of the environment inherent in vision-based approaches. However, purely IMU-based tracking compromises translation estimates and accurate relative positioning between individuals, as inertial cues are inherently self-referential and provide no direct spatial reference for others. In this paper, we present a novel approach for robustly estimating body poses and global translation for multiple individuals by leveraging the distances between sparse wearable sensors - both on each individual and across multiple individuals. Our method Group Inertial Poser estimates these absolute distances between pairs of sensors from ultra-wideband ranging (UWB) and fuses them with inertial observations as input into structured state-space models to integrate temporal motion patterns for precise 3D pose estimation. Our novel two-step optimization further leverages the estimated distances for accurately tracking people's global trajectories through the world. We also introduce GIP-DB, the first IMU+UWB dataset for two-person tracking, which comprises 200 minutes of motion recordings from 14 participants. In our evaluation, Group Inertial Poser outperforms previous state-of-the-art methods in accuracy and robustness across synthetic and real-world data, showing the promise of IMU+UWB-based multi-human motion capture in the wild. Code, models, dataset: https://github.com/eth-siplab/GroupInertialPoser

cs.CV↗
arXiv2024

An Efficient Occupancy World Model via Decoupled Dynamic Flow and Image-assisted Training

The field of autonomous driving is experiencing a surge of interest in world models, which aim to predict potential future scenarios based on historical observations. In this paper, we introduce DFIT-OccWorld, an efficient 3D occupancy world model that leverages decoupled dynamic flow and image-assisted training strategy, substantially improving 4D scene forecasting performance. To simplify the training process, we discard the previous two-stage training strategy and innovatively reformulate the occupancy forecasting problem as a decoupled voxels warping process. Our model forecasts future dynamic voxels by warping existing observations using voxel flow, whereas static voxels are easily obtained through pose transformation. Moreover, our method incorporates an image-assisted training paradigm to enhance prediction reliability. Specifically, differentiable volume rendering is adopted to generate rendered depth maps through predicted future volumes, which are adopted in render-based photometric consistency. Experiments demonstrate the effectiveness of our approach, showcasing its state-of-the-art performance on the nuScenes and OpenScene benchmarks for 4D occupancy forecasting, end-to-end motion planning and point cloud forecasting. Concretely, it achieves state-of-the-art performances compared to existing 3D world models while incurring substantially lower computational costs.

cs.CV↗
arXiv2024

D$^2$-World: An Efficient World Model through Decoupled Dynamic Flow

This technical report summarizes the second-place solution for the Predictive World Model Challenge held at the CVPR-2024 Workshop on Foundation Models for Autonomous Systems. We introduce D$^2$-World, a novel World model that effectively forecasts future point clouds through Decoupled Dynamic flow. Specifically, the past semantic occupancies are obtained via existing occupancy networks (e.g., BEVDet). Following this, the occupancy results serve as the input for a single-stage world model, generating future occupancy in a non-autoregressive manner. To further simplify the task, dynamic voxel decoupling is performed in the world model. The model generates future dynamic voxels by warping the existing observations through voxel flow, while remaining static voxels can be easily obtained through pose transformation. As a result, our approach achieves state-of-the-art performance on the OpenScene Predictive World Model benchmark, securing second place, and trains more than 300% faster than the baseline model. Code is available at https://github.com/zhanghm1995/D2-World.

cs.CV↗
arXiv2024

Benchmarking the Robustness of LiDAR Semantic Segmentation Models

When using LiDAR semantic segmentation models for safety-critical applications such as autonomous driving, it is essential to understand and improve their robustness with respect to a large range of LiDAR corruptions. In this paper, we aim to comprehensively analyze the robustness of LiDAR semantic segmentation models under various corruptions. To rigorously evaluate the robustness and generalizability of current approaches, we propose a new benchmark called SemanticKITTI-C, which features 16 out-of-domain LiDAR corruptions in three groups, namely adverse weather, measurement noise and cross-device discrepancy. Then, we systematically investigate 11 LiDAR semantic segmentation models, especially spanning different input representations (e.g., point clouds, voxels, projected images, and etc.), network architectures and training schemes. Through this study, we obtain two insights: 1) We find out that the input representation plays a crucial role in robustness. Specifically, under specific corruptions, different representations perform variously. 2) Although state-of-the-art methods on LiDAR semantic segmentation achieve promising results on clean data, they are less robust when dealing with noisy data. Finally, based on the above observations, we design a robust LiDAR segmentation model (RLSeg) which greatly boosts the robustness with simple but effective modifications. It is promising that our benchmark, comprehensive analysis, and observations can boost future research in robust LiDAR semantic segmentation for safety-critical applications.

cs.CV↗
arXiv2023

X4D-SceneFormer: Enhanced Scene Understanding on 4D Point Cloud Videos through Cross-modal Knowledge Transfer

The field of 4D point cloud understanding is rapidly developing with the goal of analyzing dynamic 3D point cloud sequences. However, it remains a challenging task due to the sparsity and lack of texture in point clouds. Moreover, the irregularity of point cloud poses a difficulty in aligning temporal information within video sequences. To address these issues, we propose a novel cross-modal knowledge transfer framework, called X4D-SceneFormer. This framework enhances 4D-Scene understanding by transferring texture priors from RGB sequences using a Transformer architecture with temporal relationship mining. Specifically, the framework is designed with a dual-branch architecture, consisting of an 4D point cloud transformer and a Gradient-aware Image Transformer (GIT). During training, we employ multiple knowledge transfer techniques, including temporal consistency losses and masked self-attention, to strengthen the knowledge transfer between modalities. This leads to enhanced performance during inference using single-modal 4D point cloud inputs. Extensive experiments demonstrate the superior performance of our framework on various 4D point cloud video understanding tasks, including action recognition, action segmentation and semantic segmentation. The results achieve 1st places, i.e., 85.3% (+7.9%) accuracy and 47.3% (+5.0%) mIoU for 4D action segmentation and semantic segmentation, on the HOI4D challenge\footnote{\url{http://www.hoi4d.top/}.}, outperforming previous state-of-the-art by a large margin. We release the code at https://github.com/jinglinglingling/X4D

cs.CV↗
arXiv2022

Uniqueness of positive solutions to elliptic equations with the critical exponential growth on the unit disc and its applications

In this paper, we will solve this uniqueness problem of positive solutions to the following equations of exponential growth: \begin{equation*} \begin{cases} -Δu =λue^{u^2},\quad\quad & x\in B_1\subset \mathbb{R}^2,\\ u>0, & x\in B_1,\ \\ u=0,\quad\quad &x\in \partial B_1, \end{cases} \end{equation*} where $ 0<λ<λ_1(B_1)$ and $λ_1(B_1)$ denotes the first eigenvalue of the operator $-Δ$ with the Dirichlet boundary in unit disk. Our method relies on delicate and difficult analysis of radial solutions to the above equation and careful asymptotic expansion of solutions near the boundary. This uniqueness result will shed some light on solving the conjecture that maximizers of the Trudinger-Moser inequality on the unit disc are unique. Furthermore, based on this uniqueness result, we develop a new strategy to establish the quantization property of elliptic equations with the critical exponential growth in the balls of hyperbolic spaces, and obtain the multiplicity and non-existence of positive critical points for super-critical Trudinger-Moser functional. Our method for the quantization property and non-existence of the critical points avoids using the complicated blow-up analysis used in the literature. This method can also be applied to study the similar problems in balls of high dimensional Euclidean space $\mathbb{R}^n$ or hyperbolic spaces provided the uniqueness for the corresponding quasilinear elliptic equations with the critical exponential growth is established.

math.AP↗
arXiv2022

Distributed Optimal Power Flow for VSC-MTDC Meshed AC/DC Grids Using ALADIN

The increasing application of voltage source converter (VSC) high voltage direct current (VSC-HVDC) technology in power grids has raised the importance of incorporating DC grids and converters into the existing transmission network. This poses significant challenges in dealing with the resulting optimal power flow (OPF) problem. In this paper, a recently proposed nonconvex distributed optimization algorithm -- Augmented Lagrangian based Alternating Direction Inexact Newton method (ALADIN), is tailored to solve the nonconvex AC/DC OPF problem for emerging voltage source converter (VSC) based multiterminal high voltage direct current (VSC-MTDC) meshed AC/DC hybrid systems. The proposed scheme decomposes this AC/DC hybrid OPF problem and handles it in a fully distributed way. Compared to the existing state-of-art Alternating Direction Method of Multipliers(ADMM), which is in general, not applicable for nonconvex problems, ALADIN has a theoretical convergence guarantee. Applying these two approaches to (VSC-MTDC) coupled with an IEEE benchmark AC power system illustrates that the tailored ALADIN outperforms ADMM in convergence speed and numerical robustness.

math.OC↗
arXiv2020

Dense Residual Network for Retinal Vessel Segmentation

Retinal vessel segmentation plays an imaportant role in the field of retinal image analysis because changes in retinal vascular structure can aid in the diagnosis of diseases such as hypertension and diabetes. In recent research, numerous successful segmentation methods for fundus images have been proposed. But for other retinal imaging modalities, more research is needed to explore vascular extraction. In this work, we propose an efficient method to segment blood vessels in Scanning Laser Ophthalmoscopy (SLO) retinal images. Inspired by U-Net, "feature map reuse" and residual learning, we propose a deep dense residual network structure called DRNet. In DRNet, feature maps of previous blocks are adaptively aggregated into subsequent layers as input, which not only facilitates spatial reconstruction, but also learns more efficiently due to more stable gradients. Furthermore, we introduce DropBlock to alleviate the overfitting problem of the network. We train and test this model on the recent SLO public dataset. The results show that our method achieves the state-of-the-art performance even without data augmentation.

eess.IV↗
arXiv2000

Hoelder Inequalities and Isospin Splitting of the Quark Scalar Mesons

A Hoelder inequality analysis of the QCD Laplace sum-rule which probes the non-strange (n\bar n) components of the I={0,1} (light-quark) scalar mesons supports the methodological consistency of an effective continuum contribution from instanton effects. This revised formulation enhances the magnitude of the instanton contributions which split the degeneracy between the I=0 and I=1 channels. Despite this enhanced isospin splitting effect, analysis of the Laplace and finite-energy sum-rules seems to preclude identification of a_0(980) and a light broad sigma-resonance state as the lightest isovector and isoscalar spin-zero $n\bar n$ mesons. This apparent decoupling of sigma [\equiv f_0(400-1200)] and a_0(980) from the quark n\bar n scalar currents suggests either a non-q \bar q or a dominantly s\bar s interpretation of these resonances, and further suggests the possible identification of the f_0(980) and a_0(1450) as the lightest I={0,1} scalar mesons containing a substantial n\bar n component.

hep-ph↗
arXiv1999

Vacuum Condensates and the Anomalous Magnetic Moment of a Dirac Fermion

We address anticipated fermion-antifermion and dimension-4 gauge-field vacuum-condensate contributions to the magnetic portion of the fermion-photon vertex function in the presence of a vacuum with nonperturbative content, such as that of QCD. We discuss how inclusion of such condensate contributions may lead to a vanishing anomalous magnetic moment, in which case vacuum condensates may account for the apparent consistency between constituent quark masses characterizing baryon magnetic moments and those characterizing baryon spectroscopy.

hep-ph↗
arXiv1997

Broad Sub-Continuum Resonances and the Case for Finite-Energy Sum-Rules

There is a need to go beyond the narrow resonance approximation for QCD sum-rule channels which are likely to exhibit sensitivity to broad resonance structures. We first discuss how the first two Laplace sum rules are altered when one goes beyond the narrow resonance approximation to include possible subcontinuum resonances with nonzero widths. We then show that the corresponding first two finite energy sum rules are insensitive to the widths of such resonances, provided their peaks are symmetric and entirely below the continuum threshold. We also discuss the reduced sensitivity of the first two finite energy sum rules to higher dimensional condensates, and show these sum rules to be insensitive to dimension > 6 condensates containing at least one q-bar q pair. We extract the direct single-instanton contribution to the F_1 sum rule for the longitudinal component of the axial-vector correlation function from the known single-instanton contribution to the lowest Laplace sum rule for the pseudoscalar channel. Finally, we demonstrate how inclusion of this instanton contribution to the finite-energy sum rule leads to both a lighter quark mass and to more phenomenologically reasonable higher-mass resonance contributions within the pseudoscalar channel.

hep-ph↗