Most Download articles

    Published in last 1 year| In last 2 years| In last 3 years| All| Most Downloaded in Recent Month | Most Downloaded in Recent Year|

    Published in last 1 year
    Please wait a minute...
    For Selected: Toggle Thumbnails
    Survey of Model Compression for Large Language Model
    GUO Jinyang, HE Changyi, YANG Ge, LIU Xianglong
    Journal of Frontiers of Computer Science and Technology    2026, 20 (1): 1-20.   DOI: 10.3778/j.issn.1673-9418.2504069
    Large language models (LLMs) have attracted considerable attention in recent years due to their strong cognitive capabilities and widespread applications in various fields. However, their tremendous demand for computation and memory makes it difficult to deploy them in resource-constrained scenarios. Model compression and acceleration techniques have thus emerged as critical approaches to reduce computational complexity and memory usage while maintaining model performance. This paper presents a comprehensive survey of recent advances in LLM compression and acceleration methods, aiming to grasp the current development status and future trends of the entire field, promote the advancement of LLM compression and acceleration technologies, and facilitate their application and implementation in both industry and academia. It begins by outlining the challenges LLMs face in terms of computational and storage overhead. Then, it categorizes and reviews the main technical approaches, including model pruning, quantization, knowledge distillation, and low-rank decomposition, highlighting their core principles, representative methods, and cutting-edge developments. In addition, the paper provides a detailed discussion on evaluation metrics such as inference latency, accuracy retention, and deployment cost, establishing a multidimensional evaluation framework. Finally, it explores the promising future directions of LLM compression methods, aiming to guide future research and industrial deployment of compressed LLMs.
    Reference | Related Articles | Metrics
    Abstract (719)
    PDF (2873)
    Survey of Deep Learning-Based UAV Single Object Tracking
    CHEN Long, SHI Lei, LI Zhihui, DING Meng, PAN Yilun
    Journal of Frontiers of Computer Science and Technology    2026, 20 (1): 40-65.   DOI: 10.3778/j.issn.1673-9418.2506046
    Deep learning-based UAV (unmanned aerial vehicle) single object tracking has emerged as a critical research area in computer vision, aiming to accurately track designated targets in aerial video sequences. UAV tracking presents unique challenges, including drastic perspective changes, variable target scales, and computational constraints. This survey systematically categorizes recent methods into three technical approaches: traditional Siamese networks, CNN-Transformer hybrid architectures, and full Transformer methods, focusing on advances from 2022 to 2025. This paper proposes innovative sub-classifcation frameworks, including: module replacement, feature post-fusion, and collaborative modeling for CNN-Transformer hybrid architectures; static computation, hybrid mechanisms, and dynamic computation for single-stream Transformer methods. These frameworks reveal the evolution from performance-oriented to efficiency-performance balanced optimization. Comprehensive evaluations on UAV123, DTB70, UAVDT, and VisDrone2018 datasets validate the advantages and limitations of different approaches. This paper identifies key challenges with future directions and engineering deployment guidance.
    Reference | Related Articles | Metrics
    Abstract (551)
    PDF (2633)
    Survey on Deep Learning Applications in Point-of-Interest Recommendation
    HUANG Ping, WANG Feng, LIU Guangteng, WU Zhongbo, LI Xiaoli, HUANG Jinzhou
    Journal of Frontiers of Computer Science and Technology    2026, 20 (3): 671-710.   DOI: 10.3778/j.issn.1673-9418.2503022
    With the proliferation of mobile devices and location-based services, massive user check-in data from location-based social networks have generated widespread attention for point-of-interest (POI) recommendation as an important location service. Addressing challenges of data sparsity, complex spatiotemporal factors, dynamic user interest changes, privacy protection, and insufficient interpretability in traditional POI recommendation methods, this paper comprehensively reviews deep learning-based POI recommendation techniques. The formal definition of POI recommendation systems is introduced, and a general framework comprising data layer, feature engineering layer, deep learning model layer, and application layer is constructed. Deep learning techniques including recurrent neural networks, long short-term memory networks, gated recurrent units, attention mechanisms, transformers, and graph neural networks are systematically analyzed for their application principles and core algorithms in POI recommendation. Mainstream dataset characteristics and evaluation metric applicability are thoroughly analyzed. Detailed classification and performance comparison are conducted for POI recommendation models based on sequence modeling, attention mechanisms, graph structures, multimodal fusion, and specific task orientations. Through practical application case analysis, the effectiveness of deep learning-driven POI recommendation systems is validated in tourist attraction recommendation, dining recommendation, urban service point recommendation, cross-city recommendation, and industrial-scale applications. Current research challenges are systematically analyzed, including technical challenges, data challenges, and application challenges, covering key issues such as computational complexity and efficiency optimization, dynamic user preference modeling, interpretability and user trust, data sparsity and cold start problems, multimodal data fusion, privacy protection and fairness. Future development trends toward computational efficiency optimization, dynamic preference modeling, intrinsic interpretability, multimodal fusion, and privacy protection are prospected.
    Reference | Related Articles | Metrics
    Abstract (240)
    PDF (2574)
    Research Survey on Emotion Recognition Large Language Models
    WU Ao, WANG Hailong, LIU Lin, SHI Wentao
    Journal of Frontiers of Computer Science and Technology    2026, 20 (3): 625-649.   DOI: 10.3778/j.issn.1673-9418.2509014
    The rapid development of large language models offers a new paradigm for emotion recognition, demonstrating significant advantages over conventional methods in complex scenario understanding, zero/few-shot generalization, and multimodal collaborative representation. This paper provides a systematic review of large-scale emotion recognition models. It first summarizes advances in unimodal emotion recognition across text, speech, visual, and physiological signals. The study then focuses on multimodal approaches, categorizing them into unified encoder architectures, hierarchical fusion architectures, and generative model architectures based on fusion strategies, and compares their design principles, key techniques, and applicable scenarios. Further analysis highlights improvements in recognition accuracy, generalization capability, and multimodal fusion, underscoring the important role and potential of large language models in this field. Finally, this paper identifies challenges such as insufficient temporal emotion modeling and cross-cultural cognitive biases, and suggests future directions including enhancing long-range temporal modeling and establishing multimodal cultural emotion benchmarks.
    Reference | Related Articles | Metrics
    Abstract (449)
    PDF (2549)
    Research Review of Deep Learning in Skin Lesions Image Segmentation
    MENG Xiangfu, LI Jiaxun, YU Chunlin, LU Yunxuan
    Journal of Frontiers of Computer Science and Technology    2026, 20 (1): 21-39.   DOI: 10.3778/j.issn.1673-9418.2504019
    Skin lesions exhibit a wide variety of types and complex clinical manifestations, ranging from benign conditions to malignant melanomas. Early detection and accurate segmentation of these lesions are critical for the diagnosis and treatment of skin cancer, particularly in the early identification and localization of high-risk lesions such as malignant melanoma, which can significantly improve patient survival rates. In recent years, deep learning techniques have achieved remarkable progress in skin lesion image segmentation, greatly enhancing both accuracy and efficiency. This paper presents a comprehensive review of deep learning research in the field of skin lesion image segmentation. First, various skin lesion imaging modalities and commonly used public datasets are introduced, along with a summary of standard evaluation metrics. Then, addressing the prevalent issues of noise and artifacts in images, a detailed discussion on various image preprocessing and augmentation techniques is provided. Subsequently, deep learning-based segmentation methods are elaborated, covering U-Net, Transformer, SAM (segment anything model), Mamba, and multi-network fusion models. The main architectural designs, advantages, limitations, and segmentation performance of these models are comparatively analyzed. Finally, the current challenges and issues in this field are examined, and future research directions are proposed, aiming to provide valuable insights for the continued development of skin lesion image segmentation.
    Reference | Related Articles | Metrics
    Abstract (339)
    PDF (2474)
    Review of Embedding Methods and Domain Alignment Techniques in Dual-Target Cross-Domain Recommendation
    HU Siyu, MEI Hongyan, YANG Haiyan, CHENG Nai, ZHANG Xiaoyu
    Journal of Frontiers of Computer Science and Technology    2026, 20 (3): 711-729.   DOI: 10.3778/j.issn.1673-9418.2504044
    As a key branch of cross-domain recommendation technology, dual-target cross-domain recommendation improves the efficiency of source domain and target domain recommendation synchronously by virtue of a two-way collaborative optimization mechanism, and is widely used in e-commerce, video distribution, news and information and other fields. This paper first introduces the domain hierarchy and overlapping scene characteristics in the cross-domain recommendation, and from the perspective of knowledge embedding methods, elaborates on the core principles of collaborative filtering embedding, graph embedding, and self-monitoring learning embedding, and compares and analyzes their technical characteristics and applicable scenarios. From the perspective of domain alignment technology, four mainstream domain alignment schemes based on feature mapping, decoupled representation learning, meta-learning and federated learning are emphatically compared, and their technical differences and practical values are summarized. Then, the main datasets and evaluation indicators in the dual-target cross-domain recommendation are systemically sorted out, and the adaptation criteria of each dataset and indicator are defined in combination with the characteristics of different cross-domain scenarios. Finally, based on the current research status and technical challenges, future development direction of dual-target cross-domain recommendation is prospected.
    Reference | Related Articles | Metrics
    Abstract (256)
    PDF (2466)
    Review of EEG-Based Multimodal Content Generation Techniques
    LIN Chengde, YANG Mingzhe, MO Chengjun, LI Guohui
    Journal of Frontiers of Computer Science and Technology    2026, 20 (2): 301-325.   DOI: 10.3778/j.issn.1673-9418.2506009
    Electroencephalogram (EEG)-based multimodal content generation is an emerging research direction in brain-computer interface (BCI) and artificial intelligence, aiming to reconstruct multimodal content by decoding EEG signals. This technology provides a novel paradigm for brain function analysis and the development of interactive systems. This review first outlines the characteristics of EEG signals and categorizes EEG-based tasks in artificial intelligence. To address inherent challenges such as low signal-to-noise ratio and limited spatial resolution, the development of preprocessing techniques is systematically reviewed, focusing on three key methods: denoising, augmentation, and super-resolution. The advantages and limitations of traditional approaches and deep learning models in improving EEG data quality and usability are compared. For feature extraction, the review comparatively analyzes traditional time-frequency-spatial methods and deep learning models, summarizes the latest technical pathways, and assesses their impact on decoding accuracy. In multimodal content generation, the research progress in image, audio, text, and video generation is covered by modality. The main network architectures and methods currently employed are analyzed, along with a discussion of the challenges faced by the technology. The current state of research and existing problems are summarized, and future development trends are presented.
    Reference | Related Articles | Metrics
    Abstract (269)
    PDF (2450)
    Visual Mamba: Structure, Practice, and Prospects
    ZHANG Xin, ZHI Min, Sarula, Arimuzha
    Journal of Frontiers of Computer Science and Technology    2026, 20 (1): 66-78.   DOI: 10.3778/j.issn.1673-9418.2503061
    Traditional convolutional neural networks (CNNs) struggle to model global features due to their limited receptive field. Though vision Transformers (ViTs) possess the advantage of sequence modeling, they face the issue of quadratic computational complexity, posing severe computational challenges for image processing. In response, researchers have begun exploring new architectures that combine efficient computation with global perception capabilities. The visual Mamba model, based on state space models (SSMs), enables global context modeling under linear computational complexity while retaining sequence modeling capabilities, marking a new stage in vision modeling based on state space models. This paper elaborates on the basic framework of the visual Mamba block, including its dual residual structure composed of residual modules, 2D selective scan (SS2D) modules, and feed-forward networks (FFN). It analyzes the working mechanisms of cross-scanning, S6 block processing, and cross-fusion within the SS2D module. The visual Mamba model is explored from three aspects: scanning methods, stacking methods, and hybrid architectures. Scanning methods include sequential scanning and dynamic scanning, with a comparative analysis of the advantages and disadvantages of different scanning strategies. Stacking methods are categorized into serial Mamba, parallel Mamba, U-shaped Mamba, and graph Mamba, with a detailed analysis on the network construction logic of each stacking structure and its adaptability in multi-scale feature extraction and long-range dependency modeling. Hybrid architectures focus on fusion forms with CNNs, Transformers, and attention mechanisms, including single-module fusion and multi-module collaborative architectures, along with an analysis of the strengths and weaknesses of each model. Through analysis, it is pointed out that the visual Mamba model overcomes the local perception limitation of CNNs and the quadratic computational complexity of Transformers. It outperforms mainstream backbone architectures in visual tasks and demonstrates tremendous potential to become a fundamental visual backbone.
    Reference | Related Articles | Metrics
    Abstract (435)
    PDF (2439)
    Review of Multi-person Abnormal Behavior Detection Based on Deep Learning
    WANG Yanjie, WANG Xiaoqiang, ZHAO Liurui, ZHUANG Xufei
    Journal of Frontiers of Computer Science and Technology    2026, 20 (2): 326-345.   DOI: 10.3778/j.issn.1673-9418.2504014
    Content with the continuous advancement of deep learning technology, abnormal behavior detection has shifted from the traditional machine learning stage to the application of deep learning methods, and the research focus of abnormal behavior detection has shifted from single-person abnormal behavior to multi-person abnormal behavior. Multi-person abnormal behavior detection based on deep learning has become a research hotspot in the field of computer vision. For multi-person abnormal behavior detection, it is necessary to select appropriate feature extraction methods and abnormal behavior detection methods according to different scenarios. In order to make researchers have a clear and systematic understanding of the existing feature extraction methods based on deep learning and abnormal behavior detection methods in multi-person scenarios, this paper systematically analyzes and summarizes the feature extraction methods based on deep learning and multi-person abnormal behavior detection methods, and looks forward to the future development direction in view of the shortcomings of the existing methods. The definition, characteristics and classification of multi-person abnormal behavior are given. According to the feature extraction method based on deep learning and the multi-person abnormal behavior detection method based on deep learning, the existing multi-person abnormal behavior detection methods are sorted out and summarized. The commonly used public abnormal behavior detection data sets are introduced, and the performance of some models on common public data sets is compared. The future research directions in this field are prospected.
    Reference | Related Articles | Metrics
    Abstract (244)
    PDF (2316)
    Advances in Graph Neural Networks for Combinatorial Optimization Problems
    ZHU Ye, DING Cangfeng, CAO Bohao, CHEN Kexin
    Journal of Frontiers of Computer Science and Technology    2026, 20 (2): 367-385.   DOI: 10.3778/j.issn.1673-9418.2502065
    Combinatorial optimization, as an important branch of mathematical optimization, focuses on finding optimal solution within a finite discrete solution space. It is widely applied in multiple fields such as computer science, mathematics, and economics. However, with the expansion of problem scales, traditional solution methods face significant challenges. In recent years, the rapid development of machine learning technologies has brought new opportunities to the research of combinatorial optimization. In particular, graph neural networks by virtue of their powerful structural modeling capabilities and feature learning advantages, have become a popular research direction for solving combinatorial optimization problems. Thus, this paper systematically studies the application of graph neural networks in various combinatorial optimization problems. Starting from the graph representation of combinatorial optimization problems, this paper comprehensively introduces various models and algorithms related to ordinary graph neural networks, bipartite graph neural networks, tripartite graph neural networks, and hypergraph neural networks. It deeply analyzes their application strategies and practical effects solving specific combinatorial optimization problems. Furthermore, this paper comprehensively summarizes existing research, providing an objective evaluation of the strengths and limitations of various methods in practical applications. Lastly, aiming at problems such as insufficient model generalization and poor interpretability when graph neural networks solve combinatorial optimization problems, it proposes possible future research directions, expecting to provide new ideas and inspirations for the further development of this field.
    Reference | Related Articles | Metrics
    Abstract (223)
    PDF (2307)
    Review of Research on Multimodal Data Fusion Methods in Meteorology
    WU Ruofei, FANG Wei, JIANG Hongru, BAO Yansong
    Journal of Frontiers of Computer Science and Technology    2026, 20 (4): 905-922.   DOI: 10.3778/j.issn.1673-9418.2508007
    Advancements in multi-source observation technologies have ushered in the era of multimodal meteorological data. However, single-modal data is inherently limited in its ability to characterize the complex and dynamic atmospheric system, failing to meet the demand for more precise predictions. Consequently, integrating multimodal data to leverage complementary information and enhance performance has become a key research frontier in meteorology. This paper addresses the core challenge of effectively integrating meteorological data from different modalities by providing a systematic review of multimodal data fusion methods. First, this paper traces the evolution of multimodal fusion techniques, with a particular focus on deep learning-based strategies. It elaborates on the core concepts, architectural features, and advantages of mainstream models, such as encoder-decoder architectures, attention mechanisms, graph neural networks, and generative adversarial networks in the context of multimodal data fusion. Then, this paper comprehensively analyzes the current applications of this technology. Drawing on a systematic compilation of publicly available multimodal meteorological datasets, this paper reviews the technology??s application in key tasks, including precipitation nowcasting and predicting the paths and intensities of typhoons and tropical cyclones. It summarizes the research progress and effectiveness of different fusion methods in these scenarios. Furthermore, this paper analyzes the key challenges currently facing multimodal fusion in meteorology. Finally, based on the identified challenges, it outlines future research directions for the field.
    Reference | Related Articles | Metrics
    Abstract (338)
    PDF (2281)
    Review of YOLO Algorithm Research for 2D Medical Image Detection
    GUO Zhen, LIU Jing, QIU Dawei, LI Yuhao
    Journal of Frontiers of Computer Science and Technology    2026, 20 (1): 79-98.   DOI: 10.3778/j.issn.1673-9418.2502055
    In recent years, the breakthrough development of artificial intelligence technology has promoted a paradigm change in the interdisciplinary field of medicine and engineering, among which the object detection algorithm based on deep learning has shown significant advantages in medical image analysis. As a typical representative of the single-stage detection framework, the YOLO (you only look once) series algorithms have demonstrated unique advantages of high real-time, strong generalization ability and precise positioning in the field of medical image analysis through the “end-to-end” detection paradigm, and have gradually become the mainstream research methods for lesion detection, cell recognition and other tasks. The research of YOLO improved algorithm for medical object detection is sorted out. Firstly, based on the dimension of algorithm architecture innovation, the core evolution path of 12 generations of basic algorithms from YOLOv1 to YOLOv11 is sorted out, and the improvement breakthroughs, advantages and limitations, and medical scene performance of each version of YOLO are compared and analyzed. Secondly, the classic open-source datasets in the field of medical object detection are summarized, and the commonly used evaluation indicators in object detection are expounded. At the same time, the literature research on the use of YOLO improved algorithm in the detection of cervical cells, blood cells, pulmonary nodules and diabetic retinopathy in 2D medical images is reviewed, and different improved methods are comprehensively compared and analyzed. Finally, this paper summarizes the medical scenarios corresponding to different improvement ideas of YOLO, and discusses the challenges and future development directions in this field.
    Reference | Related Articles | Metrics
    Abstract (354)
    PDF (2278)
    Research Status and Prospects of Omnidirectional Image Quality Assessment Based on Deep Learning
    TIAN Yingzhe, DONG Wu, LU Likun, MA Qian, ZHOU Ziyi, ZHANG Erqing
    Journal of Frontiers of Computer Science and Technology    2026, 20 (3): 650-670.   DOI: 10.3778/j.issn.1673-9418.2503014
    In recent years, with the rapid development of virtual reality technology, omnidirectional images have gradually gained widespread attention due to their significant role in providing immersive experiences. Existing objective image quality assessment methods have not effectively evaluated the quality of omnidirectional images, making it particularly necessary to design specialized objective quality assessment methods for omnidirectional images. This paper provides a comprehensive summary of the research progress in deep learning-based objective quality assessment methods for omnidirectional images. The characteristics of omnidirectional images are analyzed. Based on the input image type of the assessment methods, objective quality assessment methods for omnidirectional images are divided into four categories: equirectangular projection format quality assessment methods, segmented spherical projection format quality assessment methods, cubic projection format quality assessment methods, and viewport image quality assessment methods. The principles, characteristics, and performance of these methods are compared. The datasets and evaluation metrics used in objective omnidirectional image quality assessment are summarized. The future development directions of objective omnidirectional image quality assessment are discussed, and practical research ideas for subsequent studies are provided.
    Reference | Related Articles | Metrics
    Abstract (204)
    PDF (2210)
    Review of Image-Point Cloud Fusion Research for AGV Environment Perception
    WANG Rong'er, WU Jigang
    Journal of Frontiers of Computer Science and Technology    2026, 20 (4): 965-976.   DOI: 10.3778/j.issn.1673-9418.2504068
    With the rapid development of automated industrial production lines and intelligent logistics and warehousing systems, automated guided vehicle (AGV) serves as core component in industrial automation and smart manufacturing. The environmental perception capability constitutes a prerequisite for achieving high-precision autonomous navigation and intelligent operations. The deep integration of artificial intelligence and multimodal sensing technologies enables AGV to overcome the limitations of traditional perception modes, realizing dynamic understanding and autonomous responses in complex industrial scenarios. Image-point cloud fusion perception technology has emerged as a critical breakthrough direction for enhancing AGV perception robustness and adaptability, leveraging its advantages in 3D spatial resolution and semantic information complementarity. The evolutionary trajectory of image-point cloud fusion technology in AGV environmental perception is systematically reviewed, while performance boundaries and application scenarios of typical fusion strategies are compared, along with deep learning-based fusion methodologies being analyzed. In response to bottleneck challenges including multi-sensor calibration and long-term stability, algorithm real-time performance, conflicts between edge computing resources and algorithm complexity, and extreme condition adaptability, this paper explores technical development directions from four dimensions: unsupervised online calibration, dynamic perception-decision closed-loop optimization, algorithm-task co-design, and adaptive perception in extreme scenarios. Theoretical foundations and technical references are provided for the construction of safer, more efficient, and universally applicable AGV environmental perception systems through these proposed approaches.
    Related Articles | Metrics
    Abstract (183)
    PDF (2185)
    Recent Advances in Speech-Driven Gesture Generation
    ZHANG Yayu, WEN Yuhui, ZHANG Xinyu, JING Liping
    Journal of Frontiers of Computer Science and Technology    2026, 20 (3): 611-624.   DOI: 10.3778/j.issn.1673-9418.2505081
    In interpersonal communication, gestures enrich verbal information and facilitate information delivery. Speech-driven gesture generation aims to automatically synthesize natural, realistic, and contextually appropriate sequences of gestures conditioned on speech input. This research direction has attracted widespread attention in fields such as computer graphics and computer vision, holding significant application value in domains including film animation production, human-computer interaction, and virtual reality. Early rule-based methods suffer from inefficiency, while regression methods, despite improving generation efficiency, often result in gestures with repetitive motion patterns and limited expressiveness.?In recent years, generative models have further advanced this field, effectively enhancing the quality and diversity of generated gestures. Regarding speech-driven gesture generation methods based on generative models, this work summarizes and categorizes relevant research on generative adversarial networks, variational autoencoders, and diffusion models, analyzing their respective applications, advantages, and disadvantages in gesture generation. It further explores the controllability of speech-driven gesture generation in emotion expression, semantic consistency, and style transfer. Moreover, collaborative generation research combining facial expressions and gestures is discussed. Additionally, commonly used datasets and evaluation metrics are introduced, followed by experimental comparative analysis of representative methods. Finally, this paper concludes by summarizing the challenges in the field of speech-driven gesture generation and outlining future research trends.
    Reference | Related Articles | Metrics
    Abstract (276)
    PDF (2132)
    Advancements in Deep Learning-Based Vehicle Trajectory Prediction Research
    FANG Jinfeng, ZHANG Zhenwei, MENG Xiangfu
    Journal of Frontiers of Computer Science and Technology    2026, 20 (2): 346-366.   DOI: 10.3778/j.issn.1673-9418.2504029
    Vehicle trajectory prediction involves using artificial intelligence methods to forecast a vehicle??s future path and behavior over a given time period. In recent years, with the continuous growth in the number of vehicles, traffic-related issues have become more prevalent, making the ability to automatically perceive, understand, and predict the next route of a vehicle increasingly vital. Additionally, the widespread adoption of various traffic data collectors has led to the generation of vast amounts of vehicle trajectory data, making trajectory prediction highly valuable in several fields, including autonomous driving. This paper aims to provide a systematic review of vehicle trajectory prediction algorithms based on deep learning. First, it summarizes the key factors influencing prediction performance, such as dataset quality and driver intent. Then, traditional trajectory prediction approaches are reviewed. Building upon this foundation, this paper focuses on deep learning-based methods, including those based on recurrent neural networks, graph convolutional networks, graph attention networks, Transformers, and other deep generative models such as generative adversarial networks and variational auto-encoders. Subsequently, commonly used datasets and evaluation metrics in the field are introduced, and various deep learning methods are compared in terms of predictive performance and generalization capability. Finally, this paper discusses major challenges in vehicle trajectory prediction, such as environmental uncertainty and behavioral variability, and provides insights into potential future research directions.
    Reference | Related Articles | Metrics
    Abstract (301)
    PDF (2118)
    Survey of Model Poisoning Attack Defense Strategies in Federated Learning
    ZHANG Lei, JIANG Ge, PU Bingqian, CHANG Liang
    Journal of Frontiers of Computer Science and Technology    2026, 20 (4): 943-964.   DOI: 10.3778/j.issn.1673-9418.2504067
    Model poisoning attack is a serious threat in federated learning. In model poisoning attacks, malicious attackers inject malicious information into training data or model updates, thereby interfering with the normal convergence of the global model until their prediction results are manipulated. The concealment and diversity of model poisoning attacks make defense extremely difficult, so it has attracted extensive attention from researchers. The principle of model poisoning attack is analyzed, and the internal mechanism of how attackers destroy the performance of the global model by tampering with local training data or forging model parameters is analyzed. On this basis, this paper systematically divides the existing defense strategies into three categories: defense strategies based on malicious model analysis, which mainly identify potential malicious behaviors through model update similarity comparison and quality assessment techniques. The core of the defense strategy based on model update robust aggregation is to significantly reduce the impact of attacks by removing extreme values or automatically weighting innovations. The defense strategy based on model update encryption aggregation creatively combines the frontier technologies of differential privacy and homomorphic encryption, which greatly improves the robustness of the model while ensuring data privacy security. The advantages and disadvantages and application scenarios are analyzed and explained. The privacy protection problems and specific solutions of model poisoning attacks are analyzed in detail, and the future development direction is proposed from the perspectives of attack and defense.
    Reference | Related Articles | Metrics
    Abstract (259)
    PDF (2004)
    Review of Pedestrian Trajectory Prediction Based on Graph Neural Networks
    DU Ting, ZHUANG Xufei, WANG Yujie, LI Ziheng, LYU Jie, ZHI Yuanyuan, ZHAO Yupeng
    Journal of Frontiers of Computer Science and Technology    2026, 20 (4): 923-942.   DOI: 10.3778/j.issn.1673-9418.2507052
    Pedestrian trajectory prediction is a critical technology in scenarios such as autonomous driving and intelligent surveillance, where its accuracy directly impacts the decision-making quality of intelligent systems. Traditional methods have certain limitations in effectively modeling the complex social interactions and environmental factors influencing pedestrian movement. Graph neural networks (GNNs), with their strong relational modeling capabilities, have become an important tool in pedestrian trajectory prediction and have demonstrated significant advantages in various predictive tasks. This review systematically organizes GNN-based pedestrian trajectory prediction methods, categorizing them into three types based on model architecture: methods based on spatio-temporal graph neural networks, methods combining GNNs with generative models, and methods integrating GNNs with Transformers. The paper provides an in-depth analysis of the modeling mechanisms, advantages, and limitations of each approach, and summarizes the performance of several representative methods on public datasets. The challenges currently faced by the field are discussed, and potential future research directions are proposed.
    Reference | Related Articles | Metrics
    Abstract (303)
    PDF (1920)
    Emotion Recognition from Physiological Signals: Review of Cross-Domain Transfer and Multimodal Fusion
    SHI Pengcheng, WANG Hailong, LIU Lin
    Journal of Frontiers of Computer Science and Technology    2026, 20 (5): 1261-1278.   DOI: 10.3778/j.issn.1673-9418.2505043
    Emotion recognition based on physiological signals has gradually emerged as a frontier research direction at the intersection of biomedical engineering and psychology in recent years. It not only provides objective, quantifiable new methods for the clinical diagnosis and efficacy assessment of affective disorders such as depression and anxiety, but also demonstrates broad application prospects in scenarios involving the perception of emotional states in healthy populations, including human-computer interaction, mental health monitoring, and intelligent driving. However, domain shift issues arising from individual physiological variations, inconsistent data acquisition equipment, and contextual diversity severely constrain models?? generalization capabilities in real-world cross-domain scenarios, forming the core bottleneck hindering the technology transition from controlled laboratories to practical applications. This paper focuses on two key technical approaches to enhance generalization: cross-domain transfer learning and multimodal fusion. Regarding cross-domain transfer, the paper critically examines domain shift challenges in electroencephalogram (EEG)-based emotion recognition, systematically summarizes the methodological evolution, applicable scenarios, and performance limitations of two primary approaches: domain adaptation and domain generalization. Concerning multimodal fusion, the paper categorizes three integration paradigms: feature-level, decision-level, and hybrid fusion, while elucidating the pivotal role of EEG-peripheral signal synergistic mechanisms in enhancing model accuracy and generalization. The paper concludes with an in-depth analysis of key challenges in current research and outlines prospective directions for future breakthroughs.
    Reference | Related Articles | Metrics
    Abstract (348)
    PDF (1357)
    Survey of Graph Anomaly Detection Driven by Graph Neural Networks
    XU Dengbin, YUAN Lining, WU Peichen, LIU Zhao
    Journal of Frontiers of Computer Science and Technology    2026, 20 (5): 1207-1240.   DOI: 10.3778/j.issn.1673-9418.2505028
    Graph anomaly detection has been widely applied in network security, financial transaction monitoring, and social network analysis, aiming to identify anomalous nodes, edges, or subgraphs within graph databases. Leveraging their ability to model complex structural patterns, graph neural networks have emerged as a key approach in this domain. However, existing surveys exhibit gaps in coverage and classification coarseness. To address these shortcomings, this survey systematically reviews graph neural network-based anomaly detection research from three perspectives. It provides precise definitions of different anomaly types in graph data, and introduces a refined classification framework by grouping models according to graph data type (static vs. dynamic) and anomaly level (node, edge, subgraph, graph-level, and multi-level). It traces the development and representative architectures of various methods through the lenses of joint embedding and anomaly detection optimization, contrastive learning, and autoencoder-based reconstruction. It compiles mainstream public datasets, evaluates the performance of representative models, and delves into major challenges such as robustness, efficiency, cross-domain transferability, and lightweight deployment. Based on this analysis, the survey outlines future research directions for multi-level anomaly detection, cross-domain adaptation, and lightweight scenarios.
    Reference | Related Articles | Metrics
    Abstract (412)
    PDF (1327)
    Review of Blockchain Approaches to Countering Quantum Computing Threats
    LI Sheng, CHEN Feng, LUO Na, XING Bin, SUN Chuanheng
    Journal of Frontiers of Computer Science and Technology    2026, 20 (5): 1241-1260.   DOI: 10.3778/j.issn.1673-9418.2506030
    Amid breakthroughs in quantum computing, blockchain systems are experiencing a dual-track evolution characterized by threats from quantum attacks and empowerment through quantum technologies. Traditional blockchain architectures, which rely on public-private key cryptography and hash functions, face systemic risks to core functions such as transaction integrity verification and consensus security when exposed to quantum attacks. This paper systematically analyzes and compares two distinct technical approaches in the field of quantum-secure blockchain: post-quantum blockchain and quantum blockchain. The former enhances existing systems by integrating post-quantum cryptographic algorithms to resist quantum threats, aiming for a smooth transition from current architectures. The latter reconstructs distributed ledger technology based on quantum information properties such as entanglement, forming novel architectures involving quantum key distribution networks, quantum digital signatures, and quantum consensus algorithms. The feasibility and development potential of these approaches are examined in application scenarios such as industrial IoT. A comparative analysis is conducted on their technical constraints and implementation challenges from the perspectives of transaction efficiency, security model validation, and standardization maturity. This lays a foundation for future empirical research. Finally, potential development paths are discussed, and targeted research directions such as diversified migration and dynamic hybrid strategies are proposed to guide further studies.
    Reference | Related Articles | Metrics
    Abstract (244)
    PDF (1320)
    Review of Fatigue Driving Detection Research
    ZHAO Ya, JIANG Liuyang, JIA Di, YAO Wenda, HUANG Shiwang
    Journal of Frontiers of Computer Science and Technology    2026, 20 (5): 1279-1293.   DOI: 10.3778/j.issn.1673-9418.2509031
    Fatigue driving is a major factor contributing to traffic accidents, making fatigue detection research crucial for ensuring road traffic safety. In recent years, advances in deep learning and intelligent vehicle technologies have improved fatigue detection techniques in terms of accuracy, real-time performance, and adaptability. This paper summarizes commonly used public fatigue driving datasets and their performance evaluation metrics. It categorizes fatigue detection methods by data acquisition approach and the evolution of discrimination dimensions, ranging from subjective to objective and from indirect to direct, covering physiological signals, observable behaviors, subjective evaluations, and expert assessments, and defines relevant discrimination criteria and calculation formulas. According to signal acquisition approach, fatigue detection methods are further divided into contact-based, non-contact-based, and hybrid approaches. Representative studies from the past five years using physiological signals, facial features, driver operational behaviors, and vehicle operational states are reviewed. A tabular summary outlines each method??s classification, contact type, applied datasets, and respective strengths and limitations. Finally, this paper analyzes the challenges of deploying fatigue detection systems in real-world scenarios, including computational constraints, insufficient cross-scenario generalization, and privacy compliance, providing valuable references for understanding the research trajectory and promoting practical applications.
    Reference | Related Articles | Metrics
    Abstract (292)
    PDF (1271)
    Survey on Event Extraction: from Traditional Methods to Large Language Models
    CAO Jiangjiang, DING Cangfeng, MA Lerong, SHI Dongyan, NING Qinwei
    Journal of Frontiers of Computer Science and Technology    2026, 20 (5): 1294-1322.   DOI: 10.3778/j.issn.1673-9418.2507007
    Event extraction, as one of the core tasks in natural language processing, aims to automatically identify event elements (such as trigger words, event types, arguments, and their roles) from unstructured text and transform them into structured representations. This technology has broad application prospects and significant practical value in scenarios such as public opinion monitoring, financial analysis, and disaster early warning. Given its critical position in theoretical research and practical applications, as well as the technical challenges it faces, this paper systematically reviews the development trajectory of event extraction technology. First, it clarifies the basic definitions and structural framework of this task, analyzing the main differences in information coverage, semantic integration capabilities, and processing strategies between sentence-level and document-level event extraction. Next, from the perspective of extraction frameworks, this paper summarizes three main frameworks, including pipeline-based, joint-based, and end-to-end, and systematically analyzes their operation processes, representative methods, and applicable scenarios. Then, it outlines the technological paradigm evolution from traditional rule-based and feature-engineering methods to deep learning approaches, and further to emerging methods based on pre-trained language models and large language models, analyzing the advantages, limitations, and development potential of each method in practical applications. Meanwhile, it reviews commonly used datasets and evaluation metrics, providing standardized references for method comparison and performance evaluation. Finally, based on the current research status, this paper identifies the key challenges in event extraction, such as generalization ability, structural expression, and semantic integration, and provides an outlook on future research directions.
    Reference | Related Articles | Metrics
    Abstract (380)
    PDF (1257)
    Survey on Time-Series Data Anomaly Detection Methods
    XU Jianqiu, HU Wei, JIN Duosi, GAO Yunjun, ZHAO Xujian
    Journal of Frontiers of Computer Science and Technology    2026, 20 (6): 1517-1544.   DOI: 10.3778/j.issn.1673-9418.2510014
    Recently, time-series data anomaly detection has received widespread attention from academia and industry, with crucial applications in service monitoring, IoT (Internet of things) surveillance, medical diagnosis, and cybersecurity. Ongoing research efforts in this domain have led to a substantial accumulation of research outcomes. To this end, this paper provides a comprehensive review of anomaly detection in time-series data and proposes a unified taxonomy, which divides all methods into three categories: shallow, deep learning-based, and model selection, with the latter two categories further subdivided. This paper also offers a detailed description of the principles and characteristics for each method. Among these three categories, shallow methods do not rely on deep learning structures but instead leverage domain knowledge from outlier detection, data mining, and classical machine learning for lightweight anomaly detection. Deep learning-based methods, based on principles of prediction, reconstruction, or contrastive learning, use the representation learning capabilities of various neural network architectures to achieve anomaly detection. Model selection methods, based on model selection or model ensembling principles, integrate multiple models to achieve the domain generalization that individual models lack. Commonly used datasets, evaluation metrics, benchmarks and open-source tools for time-series data anomaly detection are analyzed and summarized. A system architecture for time-series data anomaly detection is proposed, and the complete end-to-end process from time-series data acquisition to anomaly detection application is outlined. The major challenges in time-series data anomaly detection are summarized, along with insights into future research trends.
    Reference | Related Articles | Metrics
    Abstract (437)
    PDF (908)
    Review of Deep Learning Based Lane Line Detection Methods
    HUANG Deqi, GUO Yanan, NI Zicong
    Journal of Frontiers of Computer Science and Technology    2026, 20 (6): 1545-1561.   DOI: 10.3778/j.issn.1673-9418.2508016
    With the rapid development of autonomous driving technology, lane detection, as the core task of environmental perception systems, faces key challenges such as occlusion, sudden changes in lighting, and geometric diversity in complex scenes. A systematic study is conducted on lane detection methods based on deep learning to address this issue, with a focus on analyzing the algorithm frameworks, implementation processes, and performance optimization strategies of 2D and 3D detection technologies. In terms of 2D detection, the limitations of traditional image processing methods are explored, and the applications of convolutional neural networks, fully convolutional networks, and Transformer architectures in feature extraction, instance segmentation, and lightweight design are elaborated in detail. In the aspect of 3D detection, the technical route based on LiDAR point cloud, stereo vision and multi-sensor fusion is described, and the advantages and disadvantages of different sensor combinations in geometric reconstruction and dynamic tracking are analyzed. By combining commonly used datasets and evaluation metrics, a standardized reference basis is provided for the standardized testing of algorithm performance. Analysis shows that multimodal data fusion and lightweight network design can effectively improve the stability of lane detection systems in complex scenarios such as occlusion and sudden changes in lighting, while 3D detection technology further enhances positioning accuracy by introducing spatial geometric information. Future research should focus on the construction of a universal modeling framework, innovation in multimodal complementary fusion mechanisms, enhancement of temporal reasoning capabilities in dynamic scenarios, and efficient and lightweight deployment of embedded platforms to promote the practical application of autonomous driving technology.
    Reference | Related Articles | Metrics
    Abstract (318)
    PDF (899)
    Research Progress of Graph Neural Network Using Edge Information
    ZHU Tao, GAO Guangliang, WANG Qun, XIA Lingling, LIANG Guangjun, MA Zhuo
    Journal of Frontiers of Computer Science and Technology    2026, 20 (6): 1562-1592.   DOI: 10.3778/j.issn.1673-9418.2506019
    Graph neural network is a deep learning technique that can process non-Euclidean structured complex data which can be transformed into graphs. In recent years, it has received a lot of research attention, providing a new research perspective and powerful processing tools for the study of complex high-dimensional data mining. The core problem of graph neural network technology is to learn a low dimensional and efficient embedding representation for the elements of the graph. However, due to the inertia thinking of establishing edges based on nodes in the process of graph construction, most graph neural networks are designed with nodes as the center of the model, which results in the learning of graph elements not fully utilizing the information of edges that have the same status as nodes. At present, the research on graph neural networks using edge information is still in its preliminary stage, and there is relatively little sorting and summarizing of the contributions of existing research. Given this practical situation, firstly, the problem of using edge information in graph neural networks is described and defined, and existing graph neural networks are classified into five categories from the perspective of the application of edge information. Secondly, based on the different roles played by edge information in the design of graph neural networks, the principles and structures of typical graph neural networks in each category are introduced in detail, and the structure and process of graph neural networks that embody the important way of using edge information are vividly explained using graphical analysis method. Then, the basic characteristics of graph neural networks are emphasized in important application scenarios that require the use of edge information. Finally, the future research and development of graph neural networks using edge information are discussed.
    Reference | Related Articles | Metrics
    Abstract (268)
    PDF (852)
    Review of Communication in Multi-agent Deep Reinforcement Learning from Seven Dimensions
    CHEN Rongmin, GUO Dabo, WU Hongkun, LI Chengxiang, HU Haixiao, LIU Lixiang
    Journal of Frontiers of Computer Science and Technology    2026, 20 (6): 1606-1626.   DOI: 10.3778/j.issn.1673-9418.2506028
    Deep reinforcement learning has achieved remarkable success in solving various sequential decision-making problems in single-agent machine learning. However, in real-world environments, tasks often involve multiple agents, leading to growing interest in multi-agent deep reinforcement learning (MADRL) techniques. MADRL faces challenges such as partial observability and environmental non-stationarity. Communication-based multi-agent deep reinforcement learning (CB-MADRL) addresses these issues by introducing communication mechanisms. Focusing on the communication module in CB-MADRL, and addressing the lack of systematic analysis of communication mechanisms in existing surveys, this paper classifies communication timing and proposes new classification methods for communication targets and message aggregation. A systematic seven-dimensional analytical framework is presented, encompassing communication timing, communication target, communication source, message aggregation, communication function, communication learning, and communication constraints. Based on this framework, recent CB-MADRL methods are categorized, a unified classification notation is designed, and the seven communication dimensions of each method are summarized. This paper systematically analyzes key challenges faced by the seven communication dimensions in real-world applications, including multi-modal and heterogeneous message processing, the “black-box” nature of communication protocols, and the coupling of real-world communication constraints, and proposes targeted research ideas and feasible approaches. Furthermore, it explores the interactions and synergistic optimization mechanisms among the communication dimensions, proposes a path for exploring optimal dimension combinations, and finally envisions future research directions.
    Reference | Related Articles | Metrics
    Abstract (293)
    PDF (722)
    Survey of Research on Meta-Learning Methods in Machine Translation
    HU Songlan, WANG Siriguleng, SI Qintu
    Journal of Frontiers of Computer Science and Technology    2026, 20 (6): 1593-1605.   DOI: 10.3778/j.issn.1673-9418.2508076
    Meta-learning, as a learning paradigm emphasizing cross-task generalization and rapid adaptation, has received extensive research and application in the field of machine translation in recent years. Its core principle involves enabling models to achieve swift adaptation and effective generalization in low-resource or novel task scenarios through cross-task experiential learning. Within neural machine translation, meta-learning is regarded as a potential solution pathway for addressing data scarcity, domain transfer, and linguistic consistency issues. This paper systematically reviews the research progress of meta-learning in machine translation. First, it introduces three types of classic meta-learning methods, namely metric-based, model-based, and optimization-based technical frameworks as well as their representative algorithms. Subsequently, starting from three main application scenarios, namely low-resource language translation, domain adaptation, and semantic consistency modeling, it summarizes and analyzes the technical paths and experimental performances of existing methods, revealing the universal advantages and main limitations of meta-learning in different translation tasks. Furthermore, this paper concludes three types of performance improvement principles, namely parameter optimization, representation alignment, and meta-regularization, and analyzes their complementary relationships at different task levels. Finally, it discusses the current challenges, such as multilingual representation conflicts, poor training stability, and high deployment costs, and looks forward to future research directions, including integrating large language models, optimizing edge deployment, and introducing human feedback-based meta-optimization strategies, so as to provide references for further research in this field.
    Reference | Related Articles | Metrics
    Abstract (248)
    PDF (690)
    Survey of Word Embedding Models Research
    WEN Yongqi, YANG Ruopeng, TAO Yu, ZHONG Yihao, HUANG Bo
    Journal of Frontiers of Computer Science and Technology    2026, 20 (7): 1841-1860.   DOI: 10.3778/j.issn.1673-9418.2512024
    As a foundational technology in natural language processing, word embedding models map discrete linguistic symbols into continuous vector representations that computers can process. Their representational capacity and generalization performance directly impact the effectiveness of downstream tasks. Traditional word representation methods struggle to capture semantic relationships between words, while static word embedding methods fail to handle polysemy effectively. With the development of pre-trained language models and large language models, word embedding technology has gradually evolved from fixed vector representations to dynamic representations with contextual awareness. However, current research has yet to systematically incorporate the new paradigms of embedding models in the era of large language models. This paper comprehensively reviews the development trajectory of word embedding models, dividing their evolution into four stages based on typical technical paradigms: static word embeddings based on statistics, static word embeddings based on feedforward neural networks, dynamic word embeddings based on pre-trained language models, and dynamic word embeddings based on large language models. For each stage, the representative models, core principles, and their advantages and disadvantages are elaborated. The dynamic word embedding methods based on large language models are systematically categorized into three technical approaches: pooling, prompt engineering, and fine-tuning. The intrinsic and extrinsic evaluation methods for word embedding models and their interrelationships are examined. Addressing the current limitations of word embedding models, future research directions such as multimodal fusion, efficiency optimization, low-resource adaptation, and interpretability are proposed.
    Reference | Related Articles | Metrics
    Abstract (309)
    PDF (664)
    Review of Deep Learning in Prostate Medical Image Segmentation
    LIU Yiming, LI Ming, MA Jingang, WEI Dejian
    Journal of Frontiers of Computer Science and Technology    2026, 20 (7): 1932-1953.   DOI: 10.3778/j.issn.1673-9418.2509053
    Prostate is a critical organ within the male reproductive system, with considerable influence on men??s health and overall quality of life. Accurate segmentation of prostate plays an essential role in the diagnosis, treatment planning, and prognosis evaluation of prostate cancer. In recent years, deep learning has become a core technology driving the development of prostate image segmentation, significantly improving segmentation accuracy and robustness through its powerful feature extraction and adaptive learning capabilities. This paper presents a systematic review of deep learning-based methods applied to prostate medical image segmentation. It introduces various imaging modalities and summarizes commonly used public datasets along with their key construction features and practical applicability. Additionally, the study outlines performance evaluation metrics widely adopted in prostate image segmentation tasks. The review further investigates the application and adaptation of prevalent deep learning models, including FCN, Mask R-CNN, GAN, U-Net, Transformer, SAM, Mamba, as well as integrated and hybrid architectures. A comprehensive comparison and detailed analysis of these models are provided, focusing on their network structures, functional advantages, inherent limitations, and actual segmentation performance. Finally, the review discusses current technical and clinical challenges and suggests promising research directions to facilitate further development and practical deployment in clinical settings. This paper reviews prostate segmentation from the perspective of model evolution and clinical applicability, highlighting its technical progress and clinical value.
    Reference | Related Articles | Metrics
    Abstract (213)
    PDF (582)