arXiv subjects

Xiang Wu

Publications and source records attributed to Xiang Wu.

At least 19 recordsLinked to original sources

arXiv2026

TopoCompile: Molecular Generation with Exact Global Count Constraints

Molecular generation designs novel molecules with desired structures and properties. Global count constraints specify molecular size, elemental composition, and ring content without prescribing detailed structure. Property-conditioned generators steer generation toward such targets but do not guarantee exact satisfaction. We introduce TopoCompile, which combines an autoregressive component-tree generator with a dynamic-programming compiler for supported component-additive integer counts. TopoCompile represents molecules as trees of intact ring systems and individual atoms, making each supported count additive over components. At each step, an autoregressive Transformer scores component and attachment actions, while the compiler masks actions with no target-satisfying completion. Within the fixed component vocabulary and typed attachment rules, constrained decoding guarantees exact equality for supported non-negative, component-additive integer counts. Applied at inference time, these checks let the same trained generator support different targets and count combinations without retraining. Across four benchmarks, TopoCompile achieves 100% Exact-valid yield for Common and Rare four-count target groups; the strongest baseline within each dataset reaches 44.9% or less on Rare targets. In unconditional generation, it achieves the highest validity or valid-unique-novel (V.U.N.) among the compared methods across all four benchmark datasets.

cs.LG↗

arXiv2026

Q-MAGIC: Quad-only Mesh Autoregressive Generation via Singularity-Aware Circular Relative Offset Tokenization

Autoregressive models have proven effective for triangular mesh generation by transforming it into predicting the next face, converting the non-differentiable connectivity prediction into a sequence of token predictions. However, generating quad-only meshes remains challenging because quad faces impose stricter topological constraints than triangles. The model must not only serialize quad faces into a valid token sequence, but also support unambiguous connectivity reconstruction during de-serialization. To address this, we propose Q-MAGIC, a novel autoregressive model for quad mesh generation. We first introduce a tailored serialization scheme for quad meshes, which requires at least two new vertices for the next face and is less flexible than triangles due to strict wiring constraints. To meet these requirements and reduce serialization length, we propose a singularity-aware serialization that prioritizes wiring from singular points, leading to more coherent traversals. Additionally, we propose a circular relative offset tokenization that encodes geometry mainly via local relative offsets. Built upon this representation, an autoregressive Transformer learns the underlying distribution of quad meshes via next-token prediction, enabling effective quad mesh generation. Extensive experiments demonstrate that Q-MAGIC produces more compact sequences while consistently improving geometric fidelity and mesh quality compared to state-of-the-art methods.

cs.CV↗

arXiv2026

Anomalous pressure-dependent viscosity of basaltic melts and its role in asthenosphere melt accumulation

The asthenosphere's mechanical weakness enables plate tectonics, but its origin is debated. Partial melt has been proposed to cause this softening, yet recent studies suggest that the measured viscosity minimum in basaltic melts, an essential control on melt mobility, is an experimental artifact. Using quantum mechanics-based, machine learning-accelerated molecular dynamics, we extend simulation timescales by more than a factor of 1000 and achieve percent-level precision. We show that basaltic melt exhibits a robust viscosity minimum (approximately 20% below 1-bar values) at approximately 3 GPa, driven by pressure-induced reorganization of aluminum coordination that facilitates shear relaxation while silicon-oxygen polyhedra remain structurally rigid. Our results reveal a depth-dependent rheological transition: melt mobility peaks below approximately 150 km, promoting efficient extraction, but declines sharply during ascent, causing melt to stagnate beneath the lithosphere. This mechanism provides a physical basis for the dual seismic signatures of a melt-depleted deep asthenosphere and a melt-enriched layer near the lithosphere-asthenosphere boundary.

physics.chem-ph↗

arXiv2026

Elasticity of polycrystalline davemaoite constrains its grain size in the lower mantle

Earth's lower mantle hosts large low shear-velocity provinces (LLSVPs) beneath Africa and the Pacific, but their origin remains debated. Davemaoite, a major lower-mantle mineral, has been proposed as a contributor to these anomalies; however, its experimentally measured shear modulus is approximately 30% lower than first-principles predictions, a longstanding discrepancy between theory and experiment. Here, using million-atom machine-learning molecular dynamics simulations, we show that nanoscale grain-boundary disorder, invisible to conventional x-ray diffraction, can reconcile this gap. Introducing 5.8 vol% disordered regions reduces the shear and bulk moduli by 37% and 12%, respectively. This shear-selective softening can reproduce LLSVP-like seismic anomalies in basalt-rich assemblages, but only at disorder levels corresponding to grain sizes smaller than those expected in the lower mantle. Combining the disorder-elasticity relationship with seismic constraints, we derive a minimum davemaoite grain size of approximately 100 nm. This limit is consistent with mantle grain growth models and with attenuation-based evidence for coarse-grained LLSVPs.

physics.chem-ph↗

arXiv2026

Temporal filling-in reduces attentional fluctuations in sustained visual attention

Our capacity to maintain focus on task-relevant goals over time is constrained because attentional states wax and wane moment to moment. One theoretical account posits that temporal filling-in - filling temporal blank intervals between target stimuli - reduces attentional fluctuations to enhance sustained attention. This proposal, however, lacks direct empirical validation given the difficulty of tracking attentional fluctuations via behavioral measurements. To resolve this gap, we combined a visual timing task that assessed the behavioral benefit of temporal filling-in with electroencephalographic recordings of visual steady-state evoked potentials (SSEPs), which tracked neural responses associated with periodic attentional fluctuations entrained to the rhythmic visual sequence. Behavioral improvements in sustained attention via temporal filling-in were associated with reduced attentional fluctuations, as indexed by attenuated visual SSEP amplitudes. The findings highlight a fundamental neural mechanism whereby temporal filling-in reduces attentional fluctuations to stabilize sustained visual attention over extended periods.

q-bio.NC↗

arXiv2026

Scalable Topology-Preserving Graph Coarsening: Concepts and Algorithms

Graph coarsening reduces the size of a graph while preserving certain properties. Most existing methods preserve either spectral or spatial characteristics. Recent research shows that topology-preserving coarsening methods maintain GNN performance on coarsened graphs but suffer from exponential time complexity. To address these problems, we propose Scalable Topology-Preserving Graph Coarsening (STPGC) by introducing the concepts of graph strong collapse and graph edge collapse extended from algebraic topology. STPGC comprises three new algorithms, GStrongCollapse, GEdgeCollapse, and NeighborhoodConing based on these two concepts, which eliminate dominated nodes and edges while rigorously preserving topological features. We further prove that STPGC preserves the GNN receptive field and develop approximate algorithms to accelerate GNN training. Experiments on node classification with GNNs demonstrate the efficiency and effectiveness of STPGC.

cs.LG↗

arXiv2026

Insular intracranial activity identifies multiple facial expressions via diverse, intermixed temporal patterns at the single-contact level

How neural representations in the insular cortex support emotional processing remains poorly understood, and the extent to which the insula is specialized for disgust processing remains debated. We recorded stereoelectroencephalography data from the insula while human subjects with implanted electrode contacts performed a facial emotion recognition task involving disgusted, fearful, angry, sad, neutral, and happy expressions. Expression category specificity of insular activity was assessed via pairwise comparisons of within- and between-category pattern similarities, capturing both the shape and scale of event-related potentials (ERPs) and event-related spectral perturbations (ERSPs; theta to high-gamma frequency ranges). Insular activity successfully identified all investigated expressions, mediated by diverse ERP responses intermixed across the insula. In contrast to the marked heterogeneity of insula ERP responses, the fusiform face area exhibited convergent ERP responses across expressions and contacts, with ERSPs also contributing substantially to expression identification. These findings not only elucidate the insula's neural mechanisms underlying facial emotion perception, but also establish a potential single-contact-level neural substrate for how the insula leverages its heterogeneous response profiles to act as a key hub for versatile cognitive and emotional functions.

q-bio.NC↗

arXiv2025

OptFormer: Optical Flow-Guided Attention and Phase Space Reconstruction for SST Forecasting

Sea Surface Temperature (SST) prediction plays a vital role in climate modeling and disaster forecasting. However, it remains challenging due to its nonlinear spatiotemporal dynamics and extended prediction horizons. To address this, we propose OptFormer, a novel encoder-decoder model that integrates phase-space reconstruction with a motion-aware attention mechanism guided by optical flow. Unlike conventional attention, our approach leverages inter-frame motion cues to highlight relative changes in the spatial field, allowing the model to focus on dynamic regions and capture long-range temporal dependencies more effectively. Experiments on NOAA SST datasets across multiple spatial scales demonstrate that OptFormer achieves superior performance under a 1:1 training-to-prediction setting, significantly outperforming existing baselines in accuracy and robustness.

cs.CV↗

arXiv2025

Multimodal Medical Image Classification via Synergistic Learning Pre-training

Multimodal pathological images are usually in clinical diagnosis, but computer vision-based multimodal image-assisted diagnosis faces challenges with modality fusion, especially in the absence of expert-annotated data. To achieve the modality fusion in multimodal images with label scarcity, we propose a novel ``pretraining + fine-tuning" framework for multimodal semi-supervised medical image classification. Specifically, we propose a synergistic learning pretraining framework of consistency, reconstructive, and aligned learning. By treating one modality as an augmented sample of another modality, we implement a self-supervised learning pre-train, enhancing the baseline model's feature representation capability. Then, we design a fine-tuning method for multimodal fusion. During the fine-tuning stage, we set different encoders to extract features from the original modalities and provide a multimodal fusion encoder for fusion modality. In addition, we propose a distribution shift method for multimodal fusion features, which alleviates the prediction uncertainty and overfitting risks caused by the lack of labeled samples. We conduct extensive experiments on the publicly available gastroscopy image datasets Kvasir and Kvasirv2. Quantitative and qualitative results demonstrate that the proposed method outperforms the current state-of-the-art classification methods. The code will be released at: https://github.com/LQH89757/MICS.

cs.CV↗

arXiv2025

Eu-doped CsSrCl$_3$ Large Nanocrystal Clusters with Self-Reduction Effect and Near-Unity Quantum Yield

Europium halide perovskites have emerged as promising candidates for environmental-friendly blue-emitting materials. However, their development is hindered by relative low photoluminescence quantum yields (PLQY, e.g. ~2-5% for intrinsic CsEuCl3) and poor stability against air. Here, we introduce a one-step-procedure for synthesizing Eu$^{2+}$-doped CsSrCl$_3$ large nanocrystal clusters (LNCs) with the effect of self-reduction, therefore eliminating the use of conventional reductant oleylamine (OAm) and ensuring phase purity. The CsSrCl$_3$:Eu LNCs shows photoluminescence emission centered at 430 nm with a full width at half-maximum (FWHM) of 25 nm and a PLQY of ~40%, which can be further enhanced to ~97% after passivating the surface defects by adding trioctylphosphine (TOP), the highest among all reported lead-free blue-emitting perovskite nanocrystals. The stability of CsSrCl$_3$:Eu can also be improved significantly by epitaxially growing ZnS shell on the surface. This work will shed more light on lanthanide and alkaline-earth metal (AEM)-based perovskites for nontoxic light-emitting materials.

cond-mat.mtrl-sci↗

arXiv2025

STFM: A Spatio-Temporal Information Fusion Model Based on Phase Space Reconstruction for Sea Surface Temperature Prediction

The sea surface temperature (SST), a key environmental parameter, is crucial to optimizing production planning, making its accurate prediction a vital research topic. However, the inherent nonlinearity of the marine dynamic system presents significant challenges. Current forecasting methods mainly include physics-based numerical simulations and data-driven machine learning approaches. The former, while describing SST evolution through differential equations, suffers from high computational complexity and limited applicability, whereas the latter, despite its computational benefits, requires large datasets and faces interpretability challenges. This study presents a prediction framework based solely on data-driven techniques. Using phase space reconstruction, we construct initial-delay attractor pairs with a mathematical homeomorphism and design a Spatio-Temporal Fusion Mapping (STFM) to uncover their intrinsic connections. Unlike conventional models, our method captures SST dynamics efficiently through phase space reconstruction and achieves high prediction accuracy with minimal training data in comparative tests

cs.LG↗

arXiv2025

ScaDyG:A New Paradigm for Large-scale Dynamic Graph Learning

Dynamic graphs (DGs), which capture time-evolving relationships between graph entities, have widespread real-world applications. To efficiently encode DGs for downstream tasks, most dynamic graph neural networks follow the traditional message-passing mechanism and extend it with time-based techniques. Despite their effectiveness, the growth of historical interactions introduces significant scalability issues, particularly in industry scenarios. To address this limitation, we propose ScaDyG, with the core idea of designing a time-aware scalable learning paradigm as follows: 1) Time-aware Topology Reformulation: ScaDyG first segments historical interactions into time steps (intra and inter) based on dynamic modeling, enabling weight-free and time-aware graph propagation within pre-processing. 2) Dynamic Temporal Encoding: To further achieve fine-grained graph propagation within time steps, ScaDyG integrates temporal encoding through a combination of exponential functions in a scalable manner. 3) Hypernetwork-driven Message Aggregation: After obtaining the propagated features (i.e., messages), ScaDyG utilizes hypernetwork to analyze historical dependencies, implementing node-wise representation by an adaptive temporal fusion. Extensive experiments on 12 datasets demonstrate that ScaDyG performs comparably well or even outperforms other SOTA methods in both node and link-level downstream tasks, with fewer learnable parameters and higher efficiency.

cs.LG↗

arXiv2025

SP-SLAM: Neural Real-Time Dense SLAM With Scene Priors

Neural implicit representations have recently shown promising progress in dense Simultaneous Localization And Mapping (SLAM). However, existing works have shortcomings in terms of reconstruction quality and real-time performance, mainly due to inflexible scene representation strategy without leveraging any prior information. In this paper, we introduce SP-SLAM, a novel neural RGB-D SLAM system that performs tracking and mapping in real-time. SP-SLAM computes depth images and establishes sparse voxel-encoded scene priors near the surfaces to achieve rapid convergence of the model. Subsequently, the encoding voxels computed from single-frame depth image are fused into a global volume, which facilitates high-fidelity surface reconstruction. Simultaneously, we employ tri-planes to store scene appearance information, striking a balance between achieving high-quality geometric texture mapping and minimizing memory consumption. Furthermore, in SP-SLAM, we introduce an effective optimization strategy for mapping, allowing the system to continuously optimize the poses of all historical input frames during runtime without increasing computational overhead. We conduct extensive evaluations on five benchmark datasets (Replica, ScanNet, TUM RGB-D, Synthetic RGB-D, 7-Scenes). The results demonstrate that, compared to existing methods, we achieve superior tracking accuracy and reconstruction quality, while running at a significantly faster speed.

cs.CV↗

arXiv2024

GRID: Scene-Graph-based Instruction-driven Robotic Task Planning

Recent works have shown that Large Language Models (LLMs) can facilitate the grounding of instructions for robotic task planning. Despite this progress, most existing works have primarily focused on utilizing raw images to aid LLMs in understanding environmental information. However, this approach not only limits the scope of observation but also typically necessitates extensive multimodal data collection and large-scale models. In this paper, we propose a novel approach called Graph-based Robotic Instruction Decomposer (GRID), which leverages scene graphs instead of images to perceive global scene information and iteratively plan subtasks for a given instruction. Our method encodes object attributes and relationships in graphs through an LLM and Graph Attention Networks, integrating instruction features to predict subtasks consisting of pre-defined robot actions and target objects in the scene graph. This strategy enables robots to acquire semantic knowledge widely observed in the environment from the scene graph. To train and evaluate GRID, we establish a dataset construction pipeline to generate synthetic datasets for graph-based robotic task planning. Experiments have shown that our method outperforms GPT-4 by over 25.4% in subtask accuracy and 43.6% in task accuracy. Moreover, our method achieves a real-time speed of 0.11s per inference. Experiments conducted on datasets of unseen scenes and scenes with varying numbers of objects demonstrate that the task accuracy of GRID declined by at most 3.8%, showcasing its robust cross-scene generalization ability. We validate our method in both physical simulation and the real world. More details can be found on the project page https://jackyzengl.github.io/GRID.github.io/.

cs.RO↗

arXiv2022

A Two-phase On-line Joint Scheduling for Welfare Maximization of Charging Station

The large adoption of EVs brings practical interest to the operation optimization of the charging station. The joint scheduling of pricing and charging control will achieve a win-win situation both for the charging station and EV drivers, thus enhancing the operational capability of the station. We consider this important problem in this paper and make the following contributions. First, a joint scheduling model of pricing and charging control is developed to maximize the expected social welfare of the charging station considering the Quality of Service and the price fluctuation sensitivity of EV drivers. It is formulated as a Markov decision process with variance criterion to capture uncertainties during operation. Second, a two-phase on-line policy learning algorithm is proposed to solve this joint scheduling problem. In the first phase, it implements event-based policy iteration to find the optimal pricing scheme, while in the second phase, it implements scenario-based model predictive control for smart charging under the updated pricing scheme. Third, by leveraging the performance difference theory, the optimality of the proposed algorithm is theoretically analyzed. Numerical experiments for a charging station with distributed generation and energy storage demonstrate the effectiveness of the proposed method and the improved social welfare of the charging station.

eess.SY↗

arXiv2022

Developmental Network Two, Its Optimality, and Emergent Turing Machines

Strong AI requires the learning engine to be task non-specific and to automatically construct a dynamic hierarchy of internal features. By hierarchy, we mean, e.g., short road edges and short bush edges amount to intermediate features of landmarks; but intermediate features from tree shadows are distractors that must be disregarded by the high-level landmark concept. By dynamic, we mean the automatic selection of features while disregarding distractors is not static, but instead based on dynamic statistics (e.g. because of the instability of shadows in the context of landmark). By internal features, we mean that they are not only sensory, but also motor, so that context from motor (state) integrates with sensory inputs to become a context-based logic machine. We present why strong AI is necessary for any practical AI systems that work reliably in the real world. We then present a new generation of Developmental Networks 2 (DN-2). With many new novelties beyond DN-1, the most important novelty of DN-2 is that the inhibition area of each internal neuron is neuron-specific and dynamic. This enables DN-2 to automatically construct an internal hierarchy that is fluid, whose number of areas is not static as in DN-1. To optimally use the limited resource available, we establish that DN-2 is optimal in terms of maximum likelihood, under the condition of limited learning experience and limited resources. We also present how DN-2 can learn an emergent Universal Turing Machine (UTM). Together with the optimality, we present the optimal UTM. Experiments for real-world vision-based navigation, maze planning, and audition used DN-2. They successfully showed that DN-2 is for general purposes using natural and synthetic inputs. Their automatically constructed internal representation focuses on important features while being invariant to distractors and other irrelevant context-concepts.

cs.AI↗

arXiv2021

CM-NAS: Cross-Modality Neural Architecture Search for Visible-Infrared Person Re-Identification

Visible-Infrared person re-identification (VI-ReID) aims to match cross-modality pedestrian images, breaking through the limitation of single-modality person ReID in dark environment. In order to mitigate the impact of large modality discrepancy, existing works manually design various two-stream architectures to separately learn modality-specific and modality-sharable representations. Such a manual design routine, however, highly depends on massive experiments and empirical practice, which is time consuming and labor intensive. In this paper, we systematically study the manually designed architectures, and identify that appropriately separating Batch Normalization (BN) layers is the key to bring a great boost towards cross-modality matching. Based on this observation, the essential objective is to find the optimal separation scheme for each BN layer. To this end, we propose a novel method, named Cross-Modality Neural Architecture Search (CM-NAS). It consists of a BN-oriented search space in which the standard optimization can be fulfilled subject to the cross-modality task. Equipped with the searched architecture, our method outperforms state-of-the-art counterparts in both two benchmarks, improving the Rank-1/mAP by 6.70%/6.13% on SYSU-MM01 and by 12.17%/11.23% on RegDB. Code is released at https://github.com/JDAI-CV/CM-NAS.

cs.CV↗

arXiv2021

Deep Momentum Uncertainty Hashing

Combinatorial optimization (CO) has been a hot research topic because of its theoretic and practical importance. As a classic CO problem, deep hashing aims to find an optimal code for each data from finite discrete possibilities, while the discrete nature brings a big challenge to the optimization process. Previous methods usually mitigate this challenge by binary approximation, substituting binary codes for real-values via activation functions or regularizations. However, such approximation leads to uncertainty between real-values and binary ones, degrading retrieval performance. In this paper, we propose a novel Deep Momentum Uncertainty Hashing (DMUH). It explicitly estimates the uncertainty during training and leverages the uncertainty information to guide the approximation process. Specifically, we model bit-level uncertainty via measuring the discrepancy between the output of a hashing network and that of a momentum-updated network. The discrepancy of each bit indicates the uncertainty of the hashing network to the approximate output of that bit. Meanwhile, the mean discrepancy of all bits in a hashing code can be regarded as image-level uncertainty. It embodies the uncertainty of the hashing network to the corresponding input image. The hashing bit and image with higher uncertainty are paid more attention during optimization. To the best of our knowledge, this is the first work to study the uncertainty in hashing bits. Extensive experiments are conducted on four datasets to verify the superiority of our method, including CIFAR-10, NUS-WIDE, MS-COCO, and a million-scale dataset Clothing1M. Our method achieves the best performance on all of the datasets and surpasses existing state-of-the-art methods by a large margin.

cs.CV↗