arXiv subjects

Qing Xie

Publications and source records attributed to Qing Xie.

At least 19 recordsLinked to original sources

arXiv2026

ClinFusion: A Vision-Centric Multimodal LLM System for Holistic Medical Understanding

Multimodal large language models (MLLMs) hold immense potential to revolutionize clinical practice, yet deploying them in the medical domain is fundamentally a vision-centric challenge: models must absorb knowledge from heterogeneous 2D and 3D medical images, and evaluation protocols must align with radiologists' clinical practice and provide an accurate, fine-grained and factualness-driven assessment. In this paper, we introduce ClinFusion, a vision-centric MLLM designed for holistic medical understanding that systematically addresses these limitations. We propose a compositional and cascaded vision encoder architecture featuring a Cascade Spatial-Aware Locality Fusion operator that unifies diverse 2D and native 3D medical image understanding within a fused encoder. We further introduce a vision-grounded evaluation framework, including MedIF-Bench for instruction-following assessment and a region-of-interest-grounded method for clinically aligned and factualness-driven report generation evaluation. We show that ClinFusion sets a new state-of-the-art across a comprehensive suite of 2D and 3D multimodal medical benchmarks---spanning visual question answering, report generation, and instruction following---as well as textual medical tasks, outperforming leading open-source medical MLLMs (\textit{e.g.}, Hulu-Med, Lingshu) on 20 out of 24 benchmarks and demonstrating multimodal capabilities better than powerful proprietary models such as GPT-5.2 and Gemini-3-Flash on 13 out of 16 benchmarks, and can be further augmented with agentic tool use for retrieval-augmented and tool-assisted clinical workflows. A blinded evaluation by board-certified radiologists confirms that ClinFusion produces the highest-ranked reports, and validates our RoI-grounded metric as achieving the strongest correlation with expert judgment among all automatic evaluation metrics examined.

cs.CV↗

arXiv2026

Seq103: A Unified Neuroevolution Framework for Compact Sequence Architecture Discovery

Neuroevolution is a representative neural architecture search paradigm that evolves both network topology and weights through evolutionary algorithms. In this paper, we propose Seq103, a unified NEAT-style neuroevolution framework for compact sequence architecture discovery. Seq103 consists of a shared evolutionary backbone and an optional recurrent extension. The shared backbone includes an elementary node-and-connection representation, per-class RMSE-based evaluation, mutation-based evolution with class-wise recombination, and elitism. The optional hidden-state mechanism extends the search space with hidden-state nodes and hidden connections, enabling temporal memory when step-wise recurrent inference is required. With this design, Seq103 applies the same core search pipeline to both step-wise recurrent and sample-wise feedforward sequence classification. In recurrent tasks, the hidden-state extension is enabled to provide temporal memory; in feedforward tasks, it is disabled while the shared evolutionary backbone remains unchanged. We evaluate Seq103 on 8 text classification datasets and the full UCRArchive2018 benchmark with 128 univariate time-series datasets. On step-wise tasks, Seq103 retains 86.96% of the best-baseline accuracy on average while using 34.6x to 3218.0x fewer parameters. On sample-wise tasks over the full UCRArchive2018 benchmark, Seq103 retains 81.95% of the best-baseline accuracy on average while using 11.8x to 160,601.0x fewer parameters.

cs.NE↗

arXiv2026

Progressive Prompt-Guided Cross-Modal Reasoning for Referring Image Segmentation

Referring image segmentation aims to localize and segment a target object in an image based on a free-form referring expression. The core challenge lies in effectively bridging linguistic descriptions with object-level visual representations, especially when referring expressions involve detailed attributes and complex inter-object relationships. Existing methods either rely on cross-modal alignment or employ Semantic Segmentation Prompts, but they often lack explicit reasoning mechanisms for grounding language descriptions to target regions in the image. To address these limitations, we propose PPCR, a Progressive Prompt-guided Cross-modal Reasoning framework for referring image segmentation. PPCR explicitly structures the reasoning process as a Semantic Understanding-Spatial Grounding-Instance Segmentation pipeline. Specifically, PPCR first employs multimodal large language models (MLLMs) to generate Semantic Segmentation Prompt that capture key semantic cues of the target object. Based on this semantic context, Spatial Segmentation Prompt are further generated to reason about object location and spatial extent, enabling a progressive transition from semantic understanding to spatial grounding. The Semantic and Spatial Segmentation prompts are then jointly integrated into the segmentation module to guide accurate target localization and segmentation. Extensive experiments on standard referring image segmentation benchmarks demonstrate that PPCR consistently outperforms existing methods. The code will be publicly released to facilitate reproducibility.

cs.CV↗

arXiv2026

A Sketch+Text Composed Image Retrieval Dataset for Thangka

Composed Image Retrieval (CIR) enables image retrieval by combining multiple query modalities, but existing benchmarks predominantly focus on general-domain imagery and rely on reference images with short textual modifications. As a result, they provide limited support for retrieval scenarios that require fine-grained semantic reasoning, structured visual understanding, and domain-specific knowledge. In this work, we introduce CIRThan, a sketch+text Composed Image Retrieval dataset for Thangka imagery, a culturally grounded and knowledge-specific visual domain characterized by complex structures, dense symbolic elements, and domain-dependent semantic conventions. CIRThan contains 2,287 high-quality Thangka images, each paired with a human-drawn sketch and hierarchical textual descriptions at three semantic levels, enabling composed queries that jointly express structural intent and multi-level semantic specification. We provide standardized data splits, comprehensive dataset analysis, and benchmark evaluations of representative supervised and zero-shot CIR methods. Experimental results reveal that existing CIR approaches, largely developed for general-domain imagery, struggle to effectively align sketch-based abstractions and hierarchical textual semantics with fine-grained Thangka images, particularly without in-domain supervision. We believe CIRThan offers a valuable benchmark for advancing sketch+text CIR, hierarchical semantic modeling, and multimodal retrieval in cultural heritage and other knowledge-specific visual domains. The dataset is publicly available at https://github.com/jinyuxu-whut/CIRThan.

cs.IR↗

arXiv2026

SDR-CIR: Semantic Debias Retrieval Framework for Training-Free Zero-Shot Composed Image Retrieval

Composed Image Retrieval (CIR) aims to retrieve a target image from a query composed of a reference image and modification text. Recent training-free zero-shot methods often employ Multimodal Large Language Models (MLLMs) with Chain-of-Thought (CoT) to compose a target image description for retrieval. However, due to the fuzzy matching nature of ZS-CIR, the generated description is prone to semantic bias relative to the target image. We propose SDR-CIR, a training-free Semantic Debias Ranking method based on CoT reasoning. First, Selective CoT guides the MLLM to extract visual content relevant to the modification text during image understanding, thereby reducing visual noise at the source. We then introduce a Semantic Debias Ranking with two steps, Anchor and Debias, to mitigate semantic bias. In the Anchor step, we fuse reference image features with target description features to reinforce useful semantics and supplement omitted cues. In the Debias step, we explicitly model the visual semantic contribution of the reference image to the description and incorporate it into the similarity score as a penalty term. By supplementing omitted cues while suppressing redundancy, SDR-CIR mitigates semantic bias and improves retrieval performance. Experiments on three standard CIR benchmarks show that SDR-CIR achieves state-of-the-art results among one-stage methods while maintaining high efficiency. The code is publicly available at https://github.com/suny105/SDR-CIR.

cs.IR↗

arXiv2025

KGBridge: Knowledge-Guided Prompt Learning for Non-overlapping Cross-Domain Recommendation

Knowledge Graphs (KGs), as structured knowledge bases that organize relational information across diverse domains, provide a unified semantic foundation for cross-domain recommendation (CDR). By integrating symbolic knowledge with user-item interactions, KGs enrich semantic representations, support reasoning, and enhance model interpretability. Despite this potential, existing KG-based methods still face major challenges in CDR, particularly under non-overlapping user scenarios. These challenges arise from: (C1) sensitivity to KG sparsity and popularity bias, (C2) dependence on overlapping users for domain alignment and (C3) lack of explicit disentanglement between transferable and domain-specific knowledge, which limit effective and stable knowledge transfer. To this end, we propose KGBridge, a knowledge-guided prompt learning framework for cross-domain sequential recommendation under non-overlapping user scenarios. KGBridge comprises two core components: a KG-enhanced Prompt Encoder, which models relation-level semantics as soft prompts to provide structured and dynamic priors for user sequence modeling (addressing C1), and a Two-stage Training Paradigm, which combines cross-domain pretraining and privacy-preserving fine-tuning to enable knowledge transfer without user overlap (addressing C2). By combining relation-aware semantic control with correspondence-driven disentanglement, KGBridge explicitly separates and balances domain-shared and domain-specific semantics, thereby maintaining complementarity and stabilizing adaptation during fine-tuning (addressing C3). Extensive experiments on benchmark datasets demonstrate that KGBridge consistently outperforms state-of-the-art baselines and remains robust under varying KG sparsity, highlighting its effectiveness in mitigating structural imbalance and semantic entanglement in KG-enhanced cross-domain recommendation.

cs.IR↗

arXiv2025

Enhancing Transferability and Consistency in Cross-Domain Recommendations via Supervised Disentanglement

Cross-domain recommendation (CDR) aims to alleviate the data sparsity by transferring knowledge across domains. Disentangled representation learning provides an effective solution to model complex user preferences by separating intra-domain features (domain-shared and domain-specific features), thereby enhancing robustness and interpretability. However, disentanglement-based CDR methods employing generative modeling or GNNs with contrastive objectives face two key challenges: (i) pre-separation strategies decouple features before extracting collaborative signals, disrupting intra-domain interactions and introducing noise; (ii) unsupervised disentanglement objectives lack explicit task-specific guidance, resulting in limited consistency and suboptimal alignment. To address these challenges, we propose DGCDR, a GNN-enhanced encoder-decoder framework. To handle challenge (i), DGCDR first applies GNN to extract high-order collaborative signals, providing enriched representations as a robust foundation for disentanglement. The encoder then dynamically disentangles features into domain-shared and -specific spaces, preserving collaborative information during the separation process. To handle challenge (ii), the decoder introduces an anchor-based supervision that leverages hierarchical feature relationships to enhance intra-domain consistency and cross-domain alignment. Extensive experiments on real-world datasets demonstrate that DGCDR achieves state-of-the-art performance, with improvements of up to 11.59% across key metrics. Qualitative analyses further validate its superior disentanglement quality and transferability. Our source code and datasets are available on GitHub for further comparison.

cs.IR↗

arXiv2025

LGD: Leveraging Generative Descriptions for Zero-Shot Referring Image Segmentation

Zero-shot referring image segmentation aims to locate and segment the target region based on a referring expression, with the primary challenge of aligning and matching semantics across visual and textual modalities without training. Previous works address this challenge by utilizing Vision-Language Models and mask proposal networks for region-text matching. However, this paradigm may lead to incorrect target localization due to the inherent ambiguity and diversity of free-form referring expressions. To alleviate this issue, we present LGD (Leveraging Generative Descriptions), a framework that utilizes the advanced language generation capabilities of Multi-Modal Large Language Models to enhance region-text matching performance in Vision-Language Models. Specifically, we first design two kinds of prompts, the attribute prompt and the surrounding prompt, to guide the Multi-Modal Large Language Models in generating descriptions related to the crucial attributes of the referent object and the details of surrounding objects, referred to as attribute description and surrounding description, respectively. Secondly, three visual-text matching scores are introduced to evaluate the similarity between instance-level visual features and textual features, which determines the mask most associated with the referring expression. The proposed method achieves new state-of-the-art performance on three public datasets RefCOCO, RefCOCO+ and RefCOCOg, with maximum improvements of 9.97% in oIoU and 11.29% in mIoU compared to previous methods.

cs.CV↗

arXiv2025

Digital quantum simulation of the Su-Schrieffer-Heeger model using a parameterized quantum circuit

We perform digital quantum simulations of the noninteracting Su-Schrieffer-Heeger (SSH) model using a parameterized quantum circuit. The circuit comprises two main components: the first prepares the initial state from the product state $|0\rangle^{\otimes L}$, where $L$ is the system size; the second consists of $M$ layers of brick-wall unitaries simulating time evolution. The evolution times, encoded as the rotation angles of quantum gates in the second part, are optimized variationally to minimize the energy. The SSH model exhibits two distinct topological phases, depending on the relative strengths of inter- and intra-cell hopping amplitudes. We investigate the evolution of the energy, entanglement entropy, and mutual information towards topologically trivial and nontrivial ground states. Our results find the follows: (i) When the initial and target ground states belong to the same topological phase, the variational energy decreases exponentially, the entanglement entropy quickly saturates in a system-size-independent manner, and the mutual information remains spatially localized, as the number of layers increases. (ii) When the initial and target ground states belong to different topological phases, the variational energy decreases polynomially, the entanglement entropy initially grows logarithmically before decreasing, and the mutual information spreads ballistically across the entire system, with increasing the number of layers. Furthermore, by calculating the polarization, we identify a topological phase transition occurring at an intermediate circuit layer when the initial and final target states lie in different topological characters. Finally, we experimentally confirm this topological phase transition in an 18-site system using 19 qubits on a trapped-ion quantum computer provided by Quantinuum.

quant-ph↗

arXiv2025

Representational Alignment with Chemical Induced Fit for Molecular Relational Learning

Molecular Relational Learning (MRL) is widely applied in natural sciences to predict relationships between molecular pairs by extracting structural features. The representational similarity between substructure pairs determines the functional compatibility of molecular binding sites. Nevertheless, aligning substructure representations by attention mechanisms lacks guidance from chemical knowledge, resulting in unstable model performance in chemical space (\textit{e.g.}, functional group, scaffold) shifted data. With theoretical justification, we propose the \textbf{Re}presentational \textbf{Align}ment with Chemical Induced \textbf{Fit} (ReAlignFit) to enhance the stability of MRL. ReAlignFit dynamically aligns substructure representation in MRL by introducing chemical Induced Fit-based inductive bias. In the induction process, we design the Bias Correction Function based on substructure edge reconstruction to align representations between substructure pairs by simulating chemical conformational changes (dynamic combination of substructures). ReAlignFit further integrates the Subgraph Information Bottleneck during fit process to refine and optimize substructure pairs exhibiting high chemical functional compatibility, leveraging them to generate molecular embeddings. Experimental results on nine datasets demonstrate that ReAlignFit outperforms state-of-the-art models in two tasks and significantly enhances model's stability in both rule-shifted and scaffold-shifted data distributions.

cs.LG↗

arXiv2025

Landau-Levich Scaling for Optimization of Quantum Dot Layer Morphology and Thickness for Enhanced Quantum-Dot Light-Emitting Diode Performance

Quantum dot (QD) light-emitting diodes (QLEDs) are promising candidates for next-generation displays because of their high efficiency, brightness, broad color gamut, and solution-processability. Large-scale solution-processing of electroluminescent QLEDs poses significant challenges, particularly concerning the precise control of the active layer's thickness and uniformity. These obstacles directly impact charge transport, leading to current leakage and reduced overall efficiency. Blade-coating is a prevalent and scalable solution processing technique known for its speed and minimal waste. Additionally, it allows for continuous "roll-to-roll" processing, making it highly adaptable in various applications. In this study, we demonstrate the precise control of blade speed in the Landau-Levich regime to create a uniform QD emission layer, using a commercial CdSe/ZnS QD as a representative example. QDs assemble into different morphologies on glass and the underlying layers of the QLED device due to variations in interaction energy. The QD film thickness can be modified from monolayer to multilayer by adjusting blade speed, which can be predicted by fitting the Landau-Levich-Derjaguin theory. The optimal speed at 7 mm/s results in a QD film with a surface coverage of around 163% and low roughness (1.57 nm mean square height). The QLED external quantum efficiency (EQE) of approximately 1.5% was achieved using commercially available CdSe/ZnS QDs with low photoluminescence quantum yield (PLQY), and an EQE of around 7% has been obtained using lab-made InP/ZnSe/ZnS QDs having a solution PLQY of 74%. All-blade-coated CdSe-QLEDs are further demonstrated by adopting the optimized speed for the QD layer. This method demonstrates significant potential for developing low-cost, reproducible, and scalable QLED technologies with uniform emission characteristics and low-waste production.

physics.app-ph↗

arXiv2024

Stealthy and Robust Backdoor Attack against 3D Point Clouds through Additional Point Features

Recently, 3D backdoor attacks have posed a substantial threat to 3D Deep Neural Networks (3D DNNs) designed for 3D point clouds, which are extensively deployed in various security-critical applications. Although the existing 3D backdoor attacks achieved high attack performance, they remain vulnerable to preprocessing-based defenses (e.g., outlier removal and rotation augmentation) and are prone to detection by human inspection. In pursuit of a more challenging-to-defend and stealthy 3D backdoor attack, this paper introduces the Stealthy and Robust Backdoor Attack (SRBA), which ensures robustness and stealthiness through intentional design considerations. The key insight of our attack involves applying a uniform shift to the additional point features of point clouds (e.g., reflection intensity) widely utilized as part of inputs for 3D DNNs as the trigger. Without altering the geometric information of the point clouds, our attack ensures visual consistency between poisoned and benign samples, and demonstrate robustness against preprocessing-based defenses. In addition, to automate our attack, we employ Bayesian Optimization (BO) to identify the suitable trigger. Extensive experiments suggest that SRBA achieves an attack success rate (ASR) exceeding 94% in all cases, and significantly outperforms previous SOTA methods when multiple preprocessing operations are applied during training.

cs.CV↗

arXiv2024

A Survey on Bundle Recommendation: Methods, Applications, and Challenges

In recent years, bundle recommendation systems have gained significant attention in both academia and industry due to their ability to enhance user experience and increase sales by recommending a set of items as a bundle rather than individual items. This survey provides a comprehensive review on bundle recommendation, beginning by a taxonomy for exploring product bundling. We classify it into two categories based on bundling strategy from various application domains, i.e., discriminative and generative bundle recommendation. Then we formulate the corresponding tasks of the two categories and systematically review their methods: 1) representation learning from bundle and item levels and interaction modeling for discriminative bundle recommendation; 2) representation learning from item level and bundle generation for generative bundle recommendation. Subsequently, we survey the resources of bundle recommendation including datasets and evaluation metrics, and conduct reproducibility experiments on mainstream models. Lastly, we discuss the main challenges and highlight the promising future directions in the field of bundle recommendation, aiming to serve as a useful resource for researchers and practitioners. Our code and datasets are publicly available at https://github.com/WUT-IDEA/bundle-recommendation-survey.

cs.IR↗

arXiv2024

CrimeAlarm: Towards Intensive Intent Dynamics in Fine-grained Crime Prediction

Granularity and accuracy are two crucial factors for crime event prediction. Within fine-grained event classification, multiple criminal intents may alternately exhibit in preceding sequential events, and progress differently in next. Such intensive intent dynamics makes training models hard to capture unobserved intents, and thus leads to sub-optimal generalization performance, especially in the intertwining of numerous potential events. To capture comprehensive criminal intents, this paper proposes a fine-grained sequential crime prediction framework, CrimeAlarm, that equips with a novel mutual distillation strategy inspired by curriculum learning. During the early training phase, spot-shared criminal intents are captured through high-confidence sequence samples. In the later phase, spot-specific intents are gradually learned by increasing the contribution of low-confidence sequences. Meanwhile, the output probability distributions are reciprocally learned between prediction networks to model unobserved criminal intents. Extensive experiments show that CrimeAlarm outperforms state-of-the-art methods in terms of NDCG@5, with improvements of 4.51% for the NYC16 and 7.73% for the CHI18 in accuracy measures.

cs.LG↗

arXiv2022

MealRec: A Meal Recommendation Dataset

Bundle recommendation systems aim to recommend a bundle of items for a user to consider as a whole. They have become a norm in modern life and have been applied to many real-world settings, such as product bundle recommendation, music playlist recommendation and travel package recommendation. However, compared to studies of bundle recommendation approaches in areas such as online shopping and digital music services, research on meal recommendations for restaurants in the hospitality industry has made limited progress, due largely to the lack of high-quality benchmark datasets. A publicly available dataset specialising in meal recommendation research for the research community is in urgent demand. In this paper, we introduce a meal recommendation dataset (MealRec) that aims to facilitate future research. MealRec is constructed from the user review records of Allrecipe.com, covering 1,500+ users, 7,200+ recipes and 3,800+ meals. Each recipe is described with rich information, such as ingredients, instructions, pictures, category and tags, etc; and each meal is three-course, consisting of an appetizer, a main dish and a dessert. Furthermore, we propose a category-constrained meal recommendation model that is evaluated through comparative experiments with several state-of-the-art bundle recommendation methods on MealRec. Experimental results confirm the superiority of our model and demonstrate that MealRec is a promising testbed for meal recommendation related research. The MealRec dataset and the source code of our proposed model are available at https://github.com/WUT-IDEA/MealRec for access and reproducibility.

cs.IR↗

arXiv2022

The expected values, variances and limiting distributions of Gutman index, Schultz index, multiplicative degree-Kirchhoff index and additive degree-Kirchhoff index for a class of random chain networks

There has been an upsurge of research on complex networks in recent years. The purpose of this paper is to study the mathematical properties of the random chain networks PGn with the help of graph theory. We first solve the expected value expressions of the Gutman index, Schultz index, multiplicative degree-Kirchhoff index and additive degree-Kirchhoff index, and then we get the explicit expression formulas of their variances. Finally, we find that their limiting distributions all have the probabilistic and statistical significance of normal distribution.

math.CO↗

arXiv2022

What is Next when Sequential Prediction Meets Implicitly Hard Interaction?

Hard interaction learning between source sequences and their next targets is challenging, which exists in a myriad of sequential prediction tasks. During the training process, most existing methods focus on explicitly hard interactions caused by wrong responses. However, a model might conduct correct responses by capturing a subset of learnable patterns, which results in implicitly hard interactions with some unlearned patterns. As such, its generalization performance is weakened. The problem gets more serious in sequential prediction due to the interference of substantial similar candidate targets. To this end, we propose a Hardness Aware Interaction Learning framework (HAIL) that mainly consists of two base sequential learning networks and mutual exclusivity distillation (MED). The base networks are initialized differently to learn distinctive view patterns, thus gaining different training experiences. The experiences in the form of the unlikelihood of correct responses are drawn from each other by MED, which provides mutual exclusivity knowledge to figure out implicitly hard interactions. Moreover, we deduce that the unlikelihood essentially introduces additional gradients to push the pattern learning of correct responses. Our framework can be easily extended to more peer base networks. Evaluation is conducted on four datasets covering cyber and physical spaces. The experimental results demonstrate that our framework outperforms several state-of-the-art methods in terms of top-k based metrics.

cs.LG↗

arXiv2019

New types of topological superconductors under local magnetic symmetries

We classify gapped topological superconducting (TSC) phases of one-dimensional quantum wires with local magnetic symmetries (LMSs), in which the time-reversal symmetry $\mathcal{T}$ is broken but its combinations with certain crystalline symmetry such as $M_x \mathcal{T}$, $C_{2z} \mathcal{T}$, $C_{4z}\mathcal{T}$, and $C_{6z}\mathcal{T}$ are preserved. Our results demonstrate that an equivalent BDI class TSC can be realized in the $M_x \mathcal{T}$ or $C_{2z} \mathcal{T}$ superconducting wire, which is characterized by a chiral $Z^c$ invariant. More interestingly, we also find two types of totally new TSC phases in the $C_{4z}\mathcal{T}$, and $C_{6z}\mathcal{T}$ superconducing wires, which are beyond the known AZ class, and are characterized by a helical $Z^h$ invariant and $Z^h\oplus Z^c$ invariants, respectively. In the $Z^h$ TSC phase, $Z$-pairs of MZMs are protected at each end. In the $C_{6z}\mathcal{T}$ case, the MZMs can be either chiral or helical, and even helical-chiral coexisting. The minimal models preserving $C_{4z}\mathcal{T}$ or $C_{6z}\mathcal{T}$ symmetry are presented to illustrate their novel TSC properties and MZMs.

cond-mat.supr-con↗