Selected work

Representative Papers

A Network-driven Framework for Public Event Forecasting via Dynamic Interaction Network Evolution

Aug 15, 2026

This study addresses the insufficient modeling of dynamic interaction mechanisms in public event evolution prediction by proposing the Auto-IBDLM framework. This approach models events as dynamic interaction networks, integrating network science-inspired hybrid representation learning with GRU-based temporal modules to enable automatic transformation of structural features into a compact latent space and participant growth forecasting. Experiments across thirteen real-world datasets demonstrate that the model achieves an accuracy exceeding 97%, significantly outperforming existing methods. By combining high precision, strong generalization capability, and interpretability, this work effectively enhances predictive performance for public event evolution, offering a robust solution for capturing complex dynamic interactions in evolving social systems.

0 citationsRead paper

HazeSpikeMamba: Coupling Spiking-Inspired and State-Space Features for Self-Supervised Real-World Dehazing

Aug 07, 2026

This work addresses the limited generalizability of existing image dehazing methods to real-world hazy scenes due to their reliance on synthetic data and the scarcity of real paired training samples. To overcome these challenges, we propose HazeSpikeMamba, a novel framework that integrates a spiking neural network-inspired local processing module (TPCNNSpike) with a linear-complexity state space model (an enhanced MambaIRv2) within a multi-scale U-Net architecture. Leveraging a frozen degradation network, our approach enables self-supervised domain adaptation without requiring ground-truth clear images. By uniquely combining spiking-based local coupling mechanisms with global state modeling, HazeSpikeMamba achieves state-of-the-art performance on real-world benchmarks—RTTS, URHI, and HSTS—attaining a BRISQUE score of 27.72 and NIMA of 4.87 on RTTS, while maintaining high efficiency with only 2.02M parameters and 13.27G MACs.

0 citationsRead paper

A Parameter-Specific Retrieval and Knowledge-Guided Reasoning Framework for LLM-Based GPSR Optimization in FANETs

Aug 06, 2026

This work addresses the challenge that existing GPSR-based routing protocols for flying ad hoc networks (FANETs) struggle to adaptively tune critical parameters—such as hello interval, number of multipath routes, and greedy forwarding weights—in highly dynamic environments. To overcome this limitation, the paper proposes the PMKR-GPSR framework, which innovatively integrates large language models with a knowledge-guided mechanism. Specifically, it employs parameter-specific multi-index retrieval to acquire relevant optimization experiences from historical data and constructs a constrained knowledge graph to ensure that parameter adjustments adhere to protocol semantics and feasibility constraints. Experimental results demonstrate that the proposed approach significantly improves packet delivery ratio and reduces end-to-end delay in high-mobility FANET simulations compared to baseline methods.

0 citationsRead paper

Ranking Image Fusion the Way Humans Do: A Learned Pairwise Preference Metric for Infrared-Visible Fusion Assessment

Aug 02, 2026

This work addresses the challenge that infrared and visible image fusion lacks ideal reference images, rendering existing objective metrics inadequate for accurately capturing human subjective preferences. To overcome this limitation, we propose LPIFM—the first end-to-end learned fusion evaluation model capable of handling tie judgments—by jointly modeling the infrared and visible source images alongside two candidate fused outputs to predict pairwise human preferences. We construct the first densely annotated human preference dataset spanning multiple fusion algorithms and diverse scenes, which serves as the foundation for training LPIFM. Experimental results demonstrate that our model significantly outperforms existing metrics in both pairwise judgment accuracy and ranking correlation, effectively reproducing human preference rankings consistent with the Bradley–Terry model. The code, dataset, and trained model are publicly released.

0 citationsRead paper

Fenced Citation-Context Retrieval for Case Law: Temporal Leakage and Degree Control Across Two Jurisdictions

Jul 19, 2026

This work addresses temporal leakage in legal case retrieval caused by the use of future citation contexts, which inflates performance estimates. To mitigate this, the authors propose a temporal fencing mechanism that distinguishes genuine future citations from permissible pre-query references and indexing artifacts. They introduce a temporal admission decomposition method to quantify the proportion of illusory gains attributable to future evidence. Furthermore, they design a training-free anchored channel to enable fair and deployable evaluation. On the CLERC and ECtHR-PCR benchmarks, the fenced zero-training channel achieves a 16.1-point improvement in R@1000, with only 14.9% of the gain stemming from authentic future evidence. Notably, this zero-training approach attains an R@1000 of 79.56, rivaling the strongest trained systems.

0 citationsRead paper

Recent publications

Latest Papers

A Network-driven Framework for Public Event Forecasting via Dynamic Interaction Network Evolution

Aug 15, 2026

This study addresses the insufficient modeling of dynamic interaction mechanisms in public event evolution prediction by proposing the Auto-IBDLM framework. This approach models events as dynamic interaction networks, integrating network science-inspired hybrid representation learning with GRU-based temporal modules to enable automatic transformation of structural features into a compact latent space and participant growth forecasting. Experiments across thirteen real-world datasets demonstrate that the model achieves an accuracy exceeding 97%, significantly outperforming existing methods. By combining high precision, strong generalization capability, and interpretability, this work effectively enhances predictive performance for public event evolution, offering a robust solution for capturing complex dynamic interactions in evolving social systems.

0 citationsRead paper

HazeSpikeMamba: Coupling Spiking-Inspired and State-Space Features for Self-Supervised Real-World Dehazing

Aug 07, 2026

This work addresses the limited generalizability of existing image dehazing methods to real-world hazy scenes due to their reliance on synthetic data and the scarcity of real paired training samples. To overcome these challenges, we propose HazeSpikeMamba, a novel framework that integrates a spiking neural network-inspired local processing module (TPCNNSpike) with a linear-complexity state space model (an enhanced MambaIRv2) within a multi-scale U-Net architecture. Leveraging a frozen degradation network, our approach enables self-supervised domain adaptation without requiring ground-truth clear images. By uniquely combining spiking-based local coupling mechanisms with global state modeling, HazeSpikeMamba achieves state-of-the-art performance on real-world benchmarks—RTTS, URHI, and HSTS—attaining a BRISQUE score of 27.72 and NIMA of 4.87 on RTTS, while maintaining high efficiency with only 2.02M parameters and 13.27G MACs.

0 citationsRead paper

A Parameter-Specific Retrieval and Knowledge-Guided Reasoning Framework for LLM-Based GPSR Optimization in FANETs

Aug 06, 2026

This work addresses the challenge that existing GPSR-based routing protocols for flying ad hoc networks (FANETs) struggle to adaptively tune critical parameters—such as hello interval, number of multipath routes, and greedy forwarding weights—in highly dynamic environments. To overcome this limitation, the paper proposes the PMKR-GPSR framework, which innovatively integrates large language models with a knowledge-guided mechanism. Specifically, it employs parameter-specific multi-index retrieval to acquire relevant optimization experiences from historical data and constructs a constrained knowledge graph to ensure that parameter adjustments adhere to protocol semantics and feasibility constraints. Experimental results demonstrate that the proposed approach significantly improves packet delivery ratio and reduces end-to-end delay in high-mobility FANET simulations compared to baseline methods.

0 citationsRead paper

Ranking Image Fusion the Way Humans Do: A Learned Pairwise Preference Metric for Infrared-Visible Fusion Assessment

Aug 02, 2026

This work addresses the challenge that infrared and visible image fusion lacks ideal reference images, rendering existing objective metrics inadequate for accurately capturing human subjective preferences. To overcome this limitation, we propose LPIFM—the first end-to-end learned fusion evaluation model capable of handling tie judgments—by jointly modeling the infrared and visible source images alongside two candidate fused outputs to predict pairwise human preferences. We construct the first densely annotated human preference dataset spanning multiple fusion algorithms and diverse scenes, which serves as the foundation for training LPIFM. Experimental results demonstrate that our model significantly outperforms existing metrics in both pairwise judgment accuracy and ranking correlation, effectively reproducing human preference rankings consistent with the Bradley–Terry model. The code, dataset, and trained model are publicly released.

0 citationsRead paper

Fenced Citation-Context Retrieval for Case Law: Temporal Leakage and Degree Control Across Two Jurisdictions

Jul 19, 2026

This work addresses temporal leakage in legal case retrieval caused by the use of future citation contexts, which inflates performance estimates. To mitigate this, the authors propose a temporal fencing mechanism that distinguishes genuine future citations from permissible pre-query references and indexing artifacts. They introduce a temporal admission decomposition method to quantify the proportion of illusory gains attributable to future evidence. Furthermore, they design a training-free anchored channel to enable fair and deployable evaluation. On the CLERC and ECtHR-PCR benchmarks, the fenced zero-training channel achieves a 16.1-point improvement in R@1000, with only 14.9% of the gain stemming from authentic future evidence. Notably, this zero-training approach attains an R@1000 of 79.56, rivaling the strongest trained systems.

0 citationsRead paper