Vision Transformer (ViT) has become one of the cornerstones of the computer vision field, demonstrating exceptional performance. However, its inherent high computational complexity and inference latency still pose significant obstacles for deployment in resource-constrained environments. Token pruning, by removing less informative tokens, offers an effective strategy to reduce computational overhead. However, existing pruning methods largely rely on static or local token importance scores. This myopic approach fundamentally overlooks the sequential dependency of pruning decisions and fails to capture the interaction effects between pruning decisions across layers, often neglecting the global interactions between mask variables. To address this limitation, we propose V-Pruner, a fast and globally-informed token pruning framework for Vision Transformer. V-Pruner first leverages Fisher information to perform an initial assessment of token importance, providing a principled initial prior for pruning decisions. Building on this, V-Pruner introduces a Reinforcement Learning (RL) Proximal Policy Optimization (PPO) algorithm, refining token pruning into a global sequential decision process. The algorithm combines a composite reward signal that incorporates both model performance and computational cost to guide policy exploration, effectively evaluating the long-term impact of different pruning decision combinations on global model performance. Extensive experiments on ViT-L, DeiT-B, DeiT-S, and DeiT-T demonstrate that V-Pruner achieves a better balance between accuracy, GFLOPs, inference speed, and training time, surpassing existing mainstream ViT pruning algorithms in overall performance.

Clustering is a fundamental task in graph data mining, including both node-level and graph-level clustering. While the former has been extensively explored to capture local structures and features, the latter has gained attention for its ability to capture global relationships and high-level abstractions. However, existing methods often address these two tasks in isolation, which not only wastes computational resources but also fails to fully leverage the knowledge from both levels to improve each other, hindering consistent performance improvement. To this end, we propose a novel Unified Graph Clustering Network called UGCN, which employs both local and global graph information to address node- and graph-level clustering collaboratively. In detail, we design a dual-branch projector that performs joint learning at both node and graph levels. The first branch extracts node-level features and projects them into distinct cluster layers, where the derived prototypes are used to refine graph attributes and highlight clustering-friendly substructures. In parallel, the second branch captures subgraph embeddings and aggregates them into discriminative graph-level representations. we align the two branches through joint contrastive objectives to establish a bidirectional interaction: refined prototypes guide subgraph and graph-level clustering, while graph-level pseudo-labels provide feedback to enhance node-level clustering. Extensive experimental results across seven datasets demonstrate that our method significantly outperforms existing state-of-the-art approaches.

Graph anomaly detection (GAD) has recently gained significant attention due to its broad applicability across various graph-based domains. Graph contrastive learning (GCL) has emerged as a key approach for addressing GAD challenges, particularly in the absence of labeled data. However, existing GCL methods for GAD primarily rely on data augmentation for multi-hierarchy contrastive learning, often overlooking the semantic information from high-order neighbors, which results in incomplete node representations. Furthermore, GCL-based approaches frequently prioritize node attributes while neglecting the topological structure of complex graphs. To address these limitations, we propose a novel multi-hierarchy contrastive learning graph anomaly detection with structure-awareness (MCLASt) framework. Our approach first generates two augmented views by sampling neighbors from different orders to capture rich semantic information and leverages graph convolutional networks to obtain latent node embeddings. We then introduce three hierarchical contrastive learning modules: node-node, node-subgraph, and subgraph-subgraph level, to capture multi-hierarchy consistency information across nodes with varying orders of neighbors. To further enhance feature discrimination, we incorporate a reconstruction module that preserves the essential characteristics of the original node features. Finally, we propose a multi-hierarchy anomaly inference mechanism that integrates both attribute and topological anomaly signals for more accurate anomaly detection. Extensive experiments conducted on five real-world datasets demonstrate the effectiveness and advancement of the proposed MCLASt. Our code is available at https://github.com/shaieesss/MCLASt .

Advances in speech synthesis intensify security threats, motivating real-time deepfake detection research. We investigate whether bidirectional Mamba can serve as a competitive alternative to Self-Attention in detecting synthetic speech. Our solution, Fake-Mamba, integrates an XLSR front-end with bidirectional Mamba to capture both local and global artifacts. Our core innovation introduces three efficient encoders: TransBiMamba, ConBiMamba, and PN-BiMamba. Leveraging XLSR’s rich linguistic representations, PN-BiMamba can effectively capture the subtle cues of synthetic speech. Evaluated on ASVspoof 21 LA, 21 DF, and In-The-Wild benchmarks, FakeMamba achieves 0.97 %, 1.74 %, and 5.85% EER, respectively, representing substantial relative gains over SOTA models XLSRConformer and XLSR-Mamba. The framework maintains realtime inference across utterance lengths, demonstrating strong generalization and practical viability. The code is available at https://github.com/xuanxixi/Fake-Mamba.

Pedestrian detection in intelligent transportation systems has made significant progress but faces two critical challenges: (1) insufficient fusion of complementary information between visible and infrared spectra, particularly in complex scenarios, and (2) sensitivity to illumination changes, such as low-light or overexposed conditions, leading to degraded performance. To address these issues, we propose PedDet, an adaptive spectral optimization complementarity framework which specifically enhanced and optimized for multispectral pedestrian detection. PedDet introduces the Multi-scale Spectral Feature Perception Module (MSFPM) to adaptively fuse visible and infrared features, enhancing robustness and flexibility in feature extraction. Additionally, the Illumination Robustness Feature Decoupling Module (IRFDM) improves detection stability under varying lighting by decoupling pedestrian and background features. We further design a contrastive alignment to enhance intermodal feature discrimination. Experiments on LLVIP and MSDS datasets demonstrate that PedDet achieves state-of-the-art performance, improving the mAP by 6.6 % with superior detection accuracy even in low-light conditions, marking a significant step forward for road safety.

Graph-level clustering remains a pivotal yet formidable challenge in graph learning. Recently, the integration of deep learning with representation learning has demonstrated notable advancements, yielding performance enhancements to a certain degree. However, existing methods suffer from at least one of the following issues: 1) the original graph structure has noise, and 2) during feature propagation and pooling processes, noise is gradually aggregated into the graph-level embeddings through information propagation. Consequently, these two limitations mask clustering-friendly information, leading to suboptimal graph-level clustering performance. To this end, we propose a novel Dual Boost-Driven Graph-Level Clustering Network (DBGCN) to alternately promote graph-level clustering and filtering out interference information in a unified framework. Specifically, in the pooling step, we evaluate the contribution of features at the global and optimize them using a learnable transformation matrix to obtain high-quality graph-level representation, such that the model’s reasoning capability can be improved. Moreover, to enable reliable graph-level clustering, we first identify and suppress information detrimental to clustering by evaluating similarities between graph-level representations, providing more accurate guidance for multi-view fusion. Extensive experiments demonstrated that DBGCN outperforms the state-of-the-art graph-level clustering methods on six benchmark datasets.

Graph-level clustering is a fundamental task of data mining, aiming at dividing unlabeled graphs into distinct groups. However, existing deep methods that are limited by pooling have difficulty extracting diverse and complex graph structure features, while traditional graph kernel methods rely on exhaustive substructure search, unable to adaptively handle multi-relational data. This limitation hampers producing robust and representative graph-level embeddings. To address this issue, we propose a novel Multi-Relation Graph-Kernel Strengthen Network for Graph-Level Clustering (MGSN), which integrates Multi-Relation Modeling (MRM) with graph kernel to fully employ their respective advantages. Specifically, MGSN constructs multi-relation graphs to capture diverse semantic relationships between nodes and graphs, which employ graph kernel methods to extract graph affinity, enriching the representation space. Moreover, a Relation-aware Embedding Strengthening Strategy (RESS) is designed, which adaptively aligns multi-relation information across views while strengthening graph-level features through a progressive fusion process. Extensive experiments on multiple benchmark datasets demonstrate the superiority of MGSN over state-of-the-art methods. The results highlight its ability to leverage multi-relation structures and graph kernel features, establishing a new paradigm for robust graph-level clustering.

Long-term Time Series Forecasting (LTSF) is an important task with extensive applications across diverse domains. While contemporary methodologies have achieved notable results through the integration of time and frequency domain features, significant challenges persist. Current approaches frequently disregard the information degradation inherent in Fast fourier transform (FFT) and inverse Fast fourier transform (IFFT) operations, substantially compromising predictive accuracy. Furthermore, conventional weighting mechanisms demonstrate limitations in their capacity to capture the intricate relationships between temporal and frequency representations, leading to suboptimal feature fusion and consequent information loss. To address these limitations, we present the Joint Time-Frequency Method (JTFM), a novel framework that simultaneously extracts sequence features from both temporal and frequency domains, thereby transcending single-domain constraints and enhancing feature comprehensiveness. Additionally, we introduce the Dynamic Harmonic Accumulation Weighting Mechanism (DHAWM), which surpasses traditional weighting approaches by dynamically modulating the relative contributions of temporal and frequency domain features based on sequence-specific characteristics. This adaptive mechanism strengthens the model’s feature representation capabilities and enhances forecasting precision. Empirical validation on eight real-world datasets demonstrates the JTFM’s superior performance compared to state-of-the-art baseline methods, establishing its efficacy in long-term time series forecasting applications.

Bone density prediction via CT scans to estimate T-scores is crucial, providing a more precise assessment of bone health compared to traditional methods like X-ray bone density tests, which lack spatial resolution and the ability to detect localized changes. However, CT-based prediction faces two major challenges: the high computational complexity of transformer-based architectures, which limits their deployment in portable and clinical settings, and the imbalanced, long-tailed distribution of real-world hospital data that skews predictions. To address these issues, we introduce MedConv, a convolutional model for bone density prediction that outperforms transformer models with lower computational demands. We also adapt Bal-CE loss and post-hoc logit adjustment to improve class balance. Extensive experiments on our AustinSpine dataset shows that our approach achieves up to 21% improvement in accuracy and 20% in ROC AUC over previous state-of-the-art methods. Code will be available at https://github.com/Richardqiyi/MedConv.

Although generative AI has been applied to protect wildlife in various scenarios, studies have identified the lack of an integrated application toolkit to curb illegal wildlife trade. We therefore introduce DRCO1, the first application framework that is integrated with LLM by proposing useful policies. In the Decision-Making Module, a black-box LLM, combined with a refined ReAct-like prompting template, selects policy promoters in one region. In the Restrictive-Partial-Legalization Module, our innovative Dynamic Iterative Constraint Method is employed to calculate the controlled volume of wildlife trade under the influence of policy, which may provide an idea for Explainable AI research. The Curve-fitting module fits current and future data into a curve with an 86.27% fit to the original Product Life Cycle Curve. The Optimal-Algorithm Module proposes a Dynamic WP-CUCB Algorithm with Policy Consideration to optimize the allocation of wildlife patrol resources in the region. Experiments indicate that policies generated by DRCO can lead to significant control and promote "AI for social good". Our code and more details will be open-sourced online.

Large language models (LLMs) have achieved significant success in complex tasks across various domains, but these achievements come with high computational costs and long inference delays. Pruning, as an effective optimization technique, simplifies model structures by removing redundant components, thereby improving model generalization and operational efficiency. Although existing pruning retraining-free algorithms perform excellently in pruning time, these algorithms often focus on local optimal solutions in encoder-based language models, lacking comprehensive exploration of global optimal solutions, which may affect the overall model performance. To address this issue, we propose a novel retraining-free structured pruning algorithm, named RL-Pruner. The algorithm consists of two main stages: the Mask Rearrangement Based on Asynchronous Advantage Actor-Critic (MA3C) stage and the BiConjugate Gradient Solver for Mask Tuning (BGMT) stage. It aims to explore the intra-layer interactions of mask variables and efficiently find the global optimal solution without requiring retraining. We evaluate this method using BERTBASEand DistilBERT models on the GLUE and SQuAD benchmark tests. Experimental results show that RL-Pruner significantly improves accuracy on the SQuAD1.1benchmark. Under a 60% FLOPs constraint, compared with existing pruning retraining-free algorithms, the F1 score increases by 4.25%.

Time series anomaly detection holds notable importance for risk identification and fault detection across diverse application domains. Unsupervised learning methods have become popular because they have no requirement for labels. However, due to the challenges posed by the multiplicity of abnormal patterns, the sparsity of anomalies, and the growth of data scale and complexity, these methods often fail to capture robust and representative dependencies within the time series for identifying anomalies. To enhance the ability of models to capture normal patterns of time series and avoid the retrogression of modeling ability triggered by the dependencies on high-quality prior knowledge, we propose a differencing-based contrastive representation learning framework for time series anomaly detection (DConAD). Specifically, DConAD generates differential data to provide additional information about time series and utilizes transformer-based architecture to capture spatiotemporal dependencies, which enhances the robustness of unbiased representation learning ability. Furthermore, DConAD implements a novel KL divergence-based contrastive learning paradigm that only uses positive samples to avoid deviation from reconstruction and deploys the stop-gradient strategy to compel convergence. Extensive experiments on five public datasets show the superiority and effectiveness of DConAD compared with nine baselines. The code is available at https://github.com/shaieesss/DConAD.

Fraudulent activities have significantly increased across various domains, such as e-commerce, online review platforms, and social networks, making fraud detection a critical task. Spatial Graph Neural Networks (GNNs) have been successfully applied to fraud detection tasks due to their strong inductive learning capabilities. However, existing spatial GNN-based methods often enhance the graph structure by excluding heterophilic neighbors during message passing to align with the homophilic bias of GNNs. Unfortunately, this approach can disrupt the original graph topology and increase uncertainty in predictions. To address these limitations, this paper proposes a novel framework, Dual-channel Heterophilic Message Passing (DHMP), for fraud detection. DHMP leverages a heterophily separation module to divide the graph into homophilic and heterophilic subgraphs, mitigating the low-pass inductive bias of traditional GNNs. It then applies shared weights to capture signals at different frequencies independently and incorporates a customized sampling strategy for training. This allows nodes to adaptively balance the contributions of various signals based on their labels. Extensive experiments on three real-world datasets demonstrate that DHMP outperforms existing methods, highlighting the importance of separating signals with different frequencies for improved fraud detection. The code is available at https://github.com/shaieesss/DHMP.

Remote sensing change detection tracks surface transformations over time, aiding in disaster early warning and urban change monitoring. However, current methods struggle with challenges such as complex semantic interpretation, sensor variations, seasonal changes, and suboptimal model designs, which hinder accurate detection. To overcome these issues, we propose a method combining semantic enhancement and efficient temporal cross-perception. First, we enhance relational semantics by visualizing features at various levels. Shallow features are enhanced using a reversible method, while deep features are strengthened through a graph-structured non-Euclidean approach to capture global relationships. Second, we introduce an efficient bi-temporal interaction method that uses spatial matrix compression and matrix multiplication (similar to "QKV") for cross-temporal understanding, focusing on temporal changes. Finally, to address insufficient weight learning in distant decoder layers, we apply weight normalization transfer from better-understood layers, improving understanding in the generator. Our method outperforms baseline approaches, with 3.44% IoU improvement on CLCD and 1.23% IoU improvement on Google.

Remote sensing change detection identifies and locates surface changes from multi-temporal remote sensing images, supporting applications such as environmental monitoring, disaster assessment, and land use management. Current deep learning methods for high-resolution remote sensing change detection face challenges, including high computational costs for bi-temporal interactions and inadequate utilization of feature semantics. To address these challenges, we propose a novel method that deeply considers the physical significance of bi-temporal remote sensing images for accurate change detection. Specifically, we introduce a bi-temporal cross-perception approach based on graph compression. By leveraging spatial compaction in graph structures, this method employs cross-attention-based bi-temporal interactions to enhance cross-perception and reduce pseudo-changes caused by isolated semantic understanding. Additionally, we efficiently integrate features from hierarchical CNNs by separately considering low- and high-level features. For low-level features, we apply a contrastive learning-based semantic enhancement strategy to clearly differentiate change regions from background. For high-level features, we propose an adaptive bi-temporal feature fusion method to avoid the generalization issues of fixed learnable parameters. Experimental results on the CLCD and Google datasets demonstrate that our method outperforms baseline methods in both IoU and F1 scores.

Graph Neural Networks(GNNs) play a significant role and widely applied in fraud detection tasks, exhibiting significant advancements in detection performance compared to conventional methodologies. However, within the intricate structure of fraud graphs, fraudsters usually camouflage themselves among a large number of benign entities. An effective solution to address the camouflage problem involves the incorporation of complex and abundant edge information. However, existing GNNbased methods often overlook the integration of such crucial information into the message passing process, thereby limiting their efficacy. To address the above issues, this study proposes a novel Gated Edge-augmented Graph Neural Network(GE-GNN). Our approach begins with an edge-based feature augmentation mechanism that utilizes both node and edge features within a single relation. Subsequently, we apply augmented representation to the message passing process to update the node embeddings. Furthermore, we design a gate logistic to regulate the expression of augmented information. Finally, we fuse the node features across different relations to obtain a comprehensive representation. Extensive experimental results on two real-world datasets demonstrate the proposed method achieves higher performance over several state-of-the-art methods. Our code is available at https://github.com/shaieesss/GE-GNN