Searcharxiv⌕ Search

arXiv subjects

Guoqiang Zhao

Publications and source records attributed to Guoqiang Zhao.

At least 19 recordsLinked to original sources

VoxSAOR: Semantic-Aware Occupancy Robustification via Aleatoric Uncertainty Modeling

Vision-based 3D occupancy prediction provides a dense semantic representation of scenes for autonomous driving and embodied perception, yet remains vulnerable to occlusion, depth ambiguity, and projection errors that induce spatially varying reliability in voxel features. Existing methods typically map voxel representations to semantic predictions without explicitly modeling such observation-induced feature variations. To address this limitation, we propose VoxSAOR, a semantic-aware occupancy robustification framework that models aleatoric uncertainty through feature-space proxies. It first quantifies Local Semantic Consistency (LSC) to derive a voxel-wise structural stability prior and uses cross-view semantic consensus to filter unreliable semantic correspondences. Within each semantic category, spatially matched stable and ambiguous voxels form stable-to-ambiguous feature transitions, which are modeled by a class-conditional diagonal Gaussian distribution, capturing both feature displacement and dispersion. Guided by these proxy distributions, Aleatoric Regime Ensemble Learning (AREL) trains independently parameterized prediction heads using feature perturbations at progressively increasing scales, with perturbation strength modulated by structural reliability. At inference, proxy modeling and sampling are disabled, and all heads process the same observed features; their output probabilities are then averaged to produce a deterministic consensus prediction without requiring sampling. Experiments across multiple datasets and occupancy backbones demonstrate the effectiveness of VoxSAOR in improving semantic occupancy prediction, long-tail recognition, and predictive calibration. The source code will be made publicly available at https://github.com/MengfeiD/VoxSAOR.

cs.CV↗

OmniAct3D: Leveraging Foundation Geometry and Evidence-Grounded Reasoning for Panoramic 3D Detection

Accurate 3D detection is essential for mobile embodied agents, while Vision Foundation Models (VFMs) offer transferable visual and geometric priors. Yet existing VFM-based 3D detectors rely on narrow-view monocular images or discrete perspective views, limiting coherent surround perception; equirectangular projection (ERP) instead encodes a continuous 360 scene in a single image. Direct transfer remains difficult because ERP organizes geometry and visual information differently, making object-relevant cues hard to model, localize, and preserve. We propose OmniAct3D, a framework that adapts perspective-trained VFM detectors to ERP while preserving transferable VFM priors. To resolve geometric mismatch, the ERP-Ray Geometry Adapter (ERGA-Ray) models spherical viewing rays and periodic spatial structure. To localize evidence in scene-wide context, the Visual-Action Reasoning Chain (VARC) grounds each hypothesis in relevant panoramic evidence and converts it into a structured geometric action. To recover local cues lost under fixed token budgets, the Appearance-Guided Heading Expert (AGHE) re-encodes object regions at higher resolution for heading estimation. Experiments show that OmniAct3D improves over the previous best 3D detector by 2.96 NDS points on Spheriverse and over the unadapted VFM baseline by 24.87 mAP points on PanoMMOcc. With target-specific geometry adaptation, VARC retains 95--98% of the same-configuration mAP, indicating reusable object-level 3D reasoning across sensing configurations. The source code will be made publicly available at https://github.com/FeiT-FeiTeng/OmniAct3D.

cs.CV↗

EmbPASS: Towards Cross-Embodiment Open Panoramic Segmentation

Panoramic images provide a complete 360-degree field of view, enabling comprehensive scene understanding for embodied perception. However, heterogeneous embodied platforms exhibit substantial differences in observation viewpoints and spatial layouts, giving rise to cross-embodiment observation shifts that pose additional challenges to consistent and reliable panoramic perception, while systematic studies of this problem remain limited. To bridge this gap, we introduce a new task, termed Cross-Embodiment Open Panoramic Segmentation. Meanwhile, we establish EmbPASS, a multi-platform panoramic semantic segmentation benchmark spanning Vehicle, Drone, Wearable, and Quadruped platforms under a unified semantic taxonomy, providing a testbed for systematically studying cross-embodiment panoramic perception. We further propose EPONet, an open-vocabulary panoramic semantic segmentation network that integrates Relation-Aware Metric Adapter (RAMA) and Content-Adaptive Semantic Transfer (CAST) to enhance spatial modeling and semantic transfer under heterogeneous embodied observations. Extensive experiments show that EPONet achieves the best platform-balanced performance on EmbPASS with 35.82% mIoU, outperforming the strongest baseline by 1.10%, while remaining competitive on existing panoramic segmentation benchmarks. The source code and EmbPASS benchmark will be made publicly available at https://github.com/guopj1/EmbPASS.

cs.CV↗

OmniMimic: Dynamics-completed Motion Augmentation for Multi-style Omnidirectional Quadruped Locomotion

Animal demonstrations provide quadruped robots with natural and distinctive gait styles that are difficult to specify through hand-crafted rewards. However, their narrow directional coverage leaves little style-consistent supervision for backward, lateral, and turning commands. We present OmniMimic, a training framework that turns directionally limited animal demonstrations into a single multi-gait policy over target per-axis velocity ranges. OmniMimic first combines temporal reversal, constrained dynamics completion, and sagittal reflection to construct robot-specific kinematic and physical supervision beyond the observed directions. It then expands commands progressively from the demonstrated velocity distribution toward the target per-axis bounds, and uses a shared actor with soft-gated, gait-specialized residual experts to balance reusable locomotion skills with gait-specific corrections. Across four gaits in simulation, OmniMimic reduces mean foot-position RMSE at forward and backward reference velocities by 12.9% and velocity-tracking RMSE on a uniform Cartesian command grid by 63.1%, compared with the matched APEX baseline. The project page is at https://OmniMimic.github.io.

cs.RO↗

Spheriverse: 3D Scene Understanding from Spherical Observations in the Wild

Spherical observations provide global visual context for 3D scene understanding. However, visual information is encoded in an angular domain, whereas the physical world is represented in Cartesian coordinates. This cross-space representation gap complicates geometric correspondence and semantic evidence aggregation. To delve into this challenge, we introduce Spheriverse, comprising 64,400 temporally aligned spherical image-LiDAR pairs organized into 644 sequences. The dataset spans diverse scenes, illumination, and weather conditions, with fine-grained semantic classes. We further establish benchmarks for semantic occupancy prediction, semantic mapping, and 3D object detection, evaluating 30+ methods through overall and scene-wise comparisons. For dense prediction, we propose SphereOcc, an occupancy framework that couples spherical geometry modeling with semantic evidence retrieval. Cartesian-Spherical Representation Remodeling (CSRR) incorporates spherical range-azimuth geometry into Cartesian voxel features through region-wise modulation. Spherical Evidence Re-querying (SER) then conditions queries on voxel content and range-height-azimuth geometry to adaptively retrieve relevant semantic evidence from source spherical image features. SphereOcc achieves 13.91% mIoU and 24.65% GeoIoU, yielding relative improvements of 13.9% and 9.3% over the respective best-performing methods, TPVFormer and SurroundOcc. It also ranks first in both metrics across all five scene categories, with consistent advantages across the evaluated spatial partitions and reduced fields of view. The established benchmark and source code will be available at https://feit-feiteng.github.io/Spheriverse.

cs.CV↗

O3N: Omnidirectional Open-Vocabulary Occupancy Prediction for Embodied Intelligent Robotics

The rapid evolution of consumer electronics toward embodied intelligence has accelerated the emergence of Consumer Embodied Intelligent Robotics (CEIRs), where intelligent devices are expected to perceive, understand, and interact with complex real-world environments. Understanding and reconstructing the 3D world through omnidirectional perception is therefore becoming increasingly important for CEIRs operating in complex and dynamic environments. However, existing vision-based 3D occupancy prediction methods are constrained by limited perspective inputs and a predefined training distribution, making them difficult to support embodied intelligent systems that require comprehensive and safe perception of scenes in open-world exploration. To address this, we present O3N, the first framework for open-vocabulary occupancy prediction from a single omnidirectional RGB image. O3N embeds omnidirectional voxels in a polar-spiral topology via the Polar-spiral Mamba (PsM) module, enabling continuous spatial representation and long-range context modeling across 360°. The Occupancy Cost Aggregation (OCA) module introduces a principled mechanism for unifying geometric and semantic supervision within the voxel space, ensuring consistency between reconstructed geometry and underlying semantic structure. Moreover, Natural Modality Alignment (NMA) establishes a gradient-free alignment pathway that harmonizes visual features, voxel embeddings, and text semantics, forming a consistent pixel-voxel-text representation triad for open-world perception. Extensive experiments on multiple models demonstrate that our method not only achieves state-of-the-art performance on QuadOcc and Human360Occ benchmarks but also exhibits remarkable cross-scene generalization and semantic scalability. The source code will be made publicly available at https://github.com/MengfeiD/O3N.

cs.CV↗

Panoramic Multimodal Semantic Occupancy Prediction for Quadruped Robots

Panoramic imagery provides holistic 360° visual coverage for environmental perception in quadruped robots. However, existing occupancy prediction methods are primarily designed for wheeled autonomous driving and rely heavily on RGB cues, which limits their robustness in complex, dynamically changing environments. To bridge this gap, we introduce PanoMMOcc, the first real-world panoramic multimodal occupancy dataset for quadruped robots, comprising four sensing modalities collected across diverse scenes. We further propose VoxelHound, a panoramic multimodal occupancy perception framework tailored to legged locomotion and spherical imaging. VoxelHound incorporates a Vertical Jitter Compensation (VJC) module to mitigate severe viewpoint perturbations caused by body pitch and roll during locomotion, enabling more consistent spatial reasoning, and a Multimodal Information Prompt Fusion (MIPF) module to effectively integrate panoramic visual cues with auxiliary modalities for enhanced volumetric occupancy prediction. We also establish a comprehensive benchmark on PanoMMOcc and provide detailed dataset analyses to enable systematic evaluation in challenging embodied perception scenarios. Extensive experiments demonstrate that VoxelHound achieves state-of-the-art performance on PanoMMOcc, with a +4.16 gain in mIoU. The dataset and code will be publicly released to facilitate future research on panoramic multimodal 3D perception for embodied robotic systems at https://github.com/SXDR/PanoMMOcc.

cs.RO↗

Spherical-GOF: Geometry-Aware Panoramic Gaussian Opacity Fields for 3D Scene Reconstruction

Omnidirectional images are increasingly used in robotics and vision due to their wide field of view. However, extending 3D Gaussian Splatting (3DGS) to panoramic camera models remains challenging, as existing formulations are designed for perspective projections and naive adaptations often introduce distortion and geometric inconsistencies. We present Spherical-GOF, an omnidirectional Gaussian rendering framework built upon Gaussian Opacity Fields (GOF). Unlike projection-based rasterization, Spherical-GOF performs GOF ray sampling directly on the unit sphere in spherical ray space, enabling consistent ray-Gaussian interactions for panoramic rendering. To make the spherical ray casting efficient and robust, we derive a conservative spherical bounding rule for fast ray-Gaussian culling and introduce a spherical filtering scheme that adapts Gaussian footprints to distortion-varying panoramic pixel sampling. Extensive experiments on standard panoramic benchmarks (OmniBlender and OmniPhotos) demonstrate competitive photometric quality and substantially improved geometric consistency. Compared with the strongest baseline, Spherical-GOF reduces depth reprojection error by 57% and improves cycle inlier ratio by 21%. Qualitative results show cleaner depth and more coherent normal maps, with strong robustness to global panorama rotations. We further validate generalization on OmniRob, a real-world robotic omnidirectional dataset introduced in this work, featuring UAV and quadruped platforms. The source code and the OmniRob dataset will be released at https://github.com/1170632760/Spherical-GOF.

cs.CV↗

Tilting pairs and Wakamatsu tilting pairs of subcategories over cleft extensions

Let $(\mathcal{B},\mathcal{A}, i, e, l)$ be a cleft extension of abelian categories. We prove that the functor $l$ preserves and reflects (Wakamatsu) tilting pairs of subcategories under certain conditions, unifying an abundance of known results. Then, we apply our results to the cleft extensions of module categories, and give characterizations of tilting pairs and Wakamatsu tilting pairs over $θ$-extension of rings and tensor rings, which not only recover the earlier results in this direction, but also obtain some new conclusions.

math.RT↗

Silting and tilting objects in cleft extensions of abelian categories

We establish connections between silting and tilting objects in an abelian category $\mathcal{B}$ and those in a cleft extension $\mathcal{A}$ of $\mathcal{B}$, which provides a method for constructing more silting and tilting objects. Then we apply our results to the cleft extensions of module categories, and characterize silting and tilting modules over $θ$-extension of rings. Some known results over trivial extension of rings are extended and strengthened.

math.RT↗

$\mathcal{X}$-Gorenstein projective and $\mathcal{Y}$-Gorenstein injective modules over tensor rings

Let $T_R(M)$ be a tensor ring and $\mathcal{X}$, $\mathcal{Y}$ be two classes of $R$-modules. Under certain conditions, we prove that a $T_R(M)$-module $(A, u)$ is $Ind(\mathcal{X})$-Gorenstein projective if and only if $u$ is monomorphic and $coker(u)$ is an $\mathcal{X}$-Gorenstein projective $R$-module. $\mathcal{Y}$-Gorenstein injective $T_R(M)$-modules are also explicitly described. As a consequence, the characterizations of Ding projective and Ding injective modules over $T_R(M)$ are obtained. Some applications to trivial ring extensions and Morita context rings are given.

math.RA↗

Construction of Gorenstein projective modules over tensor rings

For a tensor ring $T_R(M)$, we obtain sufficient and necessary conditions to describe all complete projective resolutions and all Gorenstein projective modules. As a consequence, we provide a method for constructing Gorenstein projective modules over $T_R(M)$ from the ones of $R$. Some applications to trivial ring extensions, Morita context rings and triangular matrix rings are given.

math.AC↗

Deep Learning-based Cross-modal Reconstruction of Vehicle Target from Sparse 3D SAR Image

Three-dimensional synthetic aperture radar (3D SAR) is an advanced active microwave imaging technology widely utilized in remote sensing area. To achieve high-resolution 3D imaging,3D SAR requires observations from multiple aspects and altitude baselines surrounding the target. However, constrained flight trajectories often lead to sparse observations, which degrade imaging quality, particularly for anisotropic man-made small targets, such as vehicles and aircraft. In the past, compressive sensing (CS) was the mainstream approach for sparse 3D SAR image reconstruction. More recently, deep learning (DL) has emerged as a powerful alternative, markedly boosting reconstruction quality and efficiency. However, existing DL-based methods typically rely solely on high-quality 3D SAR images as supervisory signals to train deep neural networks (DNNs). This unimodal learning paradigm prevents the integration of complementary information from other data modalities, which limits reconstruction performance and reduces target discriminability due to the inherent constraints of electromagnetic scattering. In this paper, we introduce cross-modal learning and propose a Cross-Modal 3D-SAR Reconstruction Network (CMAR-Net) for enhancing sparse 3D SAR images of vehicle targets by fusing optical information. Leveraging cross-modal supervision from 2D optical images and error propagation guaranteed by differentiable rendering, CMAR-Net achieves efficient training and reconstructs sparse 3D SAR images, which are derived from highly sparse-aspect observations, into visually structured 3D vehicle images. Trained exclusively on simulated data, CMAR-Net exhibits robust generalization to real-world data, outperforming state-of-the-art CS and DL methods in structural accuracy within a large-scale parking lot experiment involving numerous civilian vehicles, thereby demonstrating its strong practical applicability.

cs.CV↗

Constructions of symmetric separable equivalences and their applications

Let $Λ$ and $Γ$ be symmetrically separably equivalent Artin algebras. We prove that there exist symmetrical separable equivalences between certain endomorphism algebras of modules. As applications, we provide several methods to construct symmetrical separable equivalences from given ones and discuss when the rigidity dimension is an invariant under symmetrical separable equivalences. Moreover, we show that a symmetrical separable equivalence preserves the Frobenius-finite type, Auslander-type condition, the (strong) Nakayama conjecture, the Auslander-Gorenstein conjecture and so on.

math.RT↗

Magnetism of kagome metals $\left(\text{Fe}_{1-x} \text{Co}_{x}\right) \text{Sn}$ studied by $μ$SR

We study the magnetic properties of the metallic kagome system $\left(\mathrm{Fe}_{1-x} \mathrm{Co}_{x}\right) \mathrm{Sn}$ by a combination of Muon Spin Relaxation ($μ\mathrm{SR}$), magnetic susceptibility and Scanning Tunneling Microscopy (STM) measurements, in single crystal specimens with Co concentrations $\mathrm{x}=0,0.11,0.8$. In the undoped antiferromagnetic compound FeSn, we find possible signatures for a previously unidentified phase that sets in at $T^*\sim 50$ K, well beneath the Neel temperature $T_N \sim 376$ K, as indicated by a peak in the relaxation rate $1/T_1$ observed in zero field (ZF) and longitudinal field (LF) $μ\mathrm{SR}$ measurements, with a corresponding anomaly in the ac and dc-susceptibility, and an increase in the static width $1/T_2$ in ZF measurements. No signatures of spatial symmetry breaking are found in STM down to $7$ K. In $\mathrm{Fe}_{0.2} \mathrm{Co}_{0.8} \mathrm{Sn}$, we find canonical spin glass behavior with freezing temperature $T_{g} \sim 3.5 \mathrm{~K}$; the ZF and LF time spectra exhibit results similar to those observed in dilute alloy spin glasses CuMn and AuFe, with a critical behavior of $1 / T_{1}$ at $T_{g}$ and $1 / \mathrm{T}_{1}\rightarrow 0$ as $T \rightarrow 0$. The absence of spin dynamics at low temperatures makes a clear contrast to the spin dynamics observed by $μ\mathrm{SR}$ in many geometrically frustrated spin systems on insulating kagome, pyrochlore, and triangular lattices. The spin glass behavior of CoSn doped with dilute Fe moments is shown to originate primarily from the randomness of doped Fe moments rather than due to geometrical frustration of the underlying lattice.

cond-mat.str-el↗

Unveiling the Potential of Segment Anything Model 2 for RGB-Thermal Semantic Segmentation with Language Guidance

The perception capability of robotic systems relies on the richness of the dataset. Although Segment Anything Model 2 (SAM2), trained on large datasets, demonstrates strong perception potential in perception tasks, its inherent training paradigm prevents it from being suitable for RGB-T tasks. To address these challenges, we propose SHIFNet, a novel SAM2-driven Hybrid Interaction Paradigm that unlocks the potential of SAM2 with linguistic guidance for efficient RGB-Thermal perception. Our framework consists of two key components: (1) Semantic-Aware Cross-modal Fusion (SACF) module that dynamically balances modality contributions through text-guided affinity learning, overcoming SAM2's inherent RGB bias; (2) Heterogeneous Prompting Decoder (HPD) that enhances global semantic information through a semantic enhancement module and then combined with category embeddings to amplify cross-modal semantic consistency. With 32.27M trainable parameters, SHIFNet achieves state-of-the-art segmentation performance on public benchmarks, reaching 89.8% on PST900 and 67.8% on FMB, respectively. The framework facilitates the adaptation of pre-trained large models to RGB-T segmentation tasks, effectively mitigating the high costs associated with data collection while endowing robotic systems with comprehensive perception capabilities. The source code will be made publicly available at https://github.com/iAsakiT3T/SHIFNet.

cs.CV↗

Gorenstein categories and separable equivalences

Let $\mathscr{C}$ be an additive subcategory of left $Λ$-modules, we establish relations of the orthogonal classes of $\mathscr{C}$ and (co)res $\widetilde{\mathscr{C}}$ under separable equivalences. As applications, we obtain that the (one-sided) Gorenstein category and Wakamatsu tilting module are preserved under separable equivalences. Furthermore, we discuss when $G_{C}$-projective (injective) modules and Auslander (Bass) class with respect to $C$ are invariant under separable equivalences.

math.RT↗

$ω$-left approximation dimensions under Stable equivalence

In this paper, we investigate some transfer properties of $ω$-left approximation dimensions of modules of stably equivalent Artin algebras having neither nodes nor semisimple direct summands. As applications, we give a one-to-one correspondence between basic (Wakamatsu) tilting modules, and prove that the Wakamatsu tilting conjecture is preserved under those equivalences.

math.RT↗