IEEE Transactions on Circuits and Systems for Video Technology最新文献_第7页

Learning Dynamic View Reliability for Multi-Camera Multi-Object Tracking 多相机多目标跟踪的动态视图可靠性学习

IF 10.8 1区工程技术

IEEE Transactions on Circuits and Systems for Video Technology Pub Date : 2026-10-01 Epub Date: 2026-06-25 DOI: 10.1109/TCSVT.2026.3707405

Song Tang;Yue Yang;Qijun Zhao

{"title":"Learning Dynamic View Reliability for Multi-Camera Multi-Object Tracking","authors":"Song Tang;Yue Yang;Qijun Zhao","doi":"10.1109/TCSVT.2026.3707405","DOIUrl":"https://doi.org/10.1109/TCSVT.2026.3707405","url":null,"abstract":"Multi-Camera Multi-Object Tracking (MC-MOT) aims to maintain consistent identities across views, yet it remains challenging when visual evidence is corrupted by view-dependent occlusions and varying visibility. While recent graph-based methods offer flexible spatio-temporal modeling, most rely on view-invariant fusion strategies that treat all camera inputs as equally reliable. This “blind” aggregation allows noise from occluded views to propagate into the target representation, leading to frequent identity switches. To address this limitation, we propose VistaTrack, a hierarchical graph-based framework that explicitly models the time-varying reliability of camera views. Central to our approach is a novel Dynamic Feature Fusion (DFF) module inserted between spatial clustering and temporal linking. By employing a dual-stream robust consensus mechanism, DFF dynamically estimates instance-level reliability scores, effectively suppressing noisy observations while emphasizing informative viewpoints. The resulting reliability-weighted identity embeddings are then associated across time to form coherent trajectories. Extensive experiments on the WildTrack, CAMPUS, and EPFL benchmarks demonstrate that VistaTrack achieves new state-of-the-art performance, showing significant improvements in IDF1 and tracking stability under severe occlusions.","PeriodicalId":13082,"journal":{"name":"IEEE Transactions on Circuits and Systems for Video Technology","volume":"36 10","pages":"16036-16050"},"PeriodicalIF":10.8,"publicationDate":"2026-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149097762","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

SAHs: Unraveling Safety Attention Heads in Large Vision-Language Models 在大视觉语言模型中揭示安全注意头

IF 10.8 1区工程技术

IEEE Transactions on Circuits and Systems for Video Technology Pub Date : 2026-10-01 Epub Date: 2026-07-07 DOI: 10.1109/TCSVT.2026.3710895

Ziwei Zheng;Lijun He;Junyao Zhao;Le Yang;Fan Li

{"title":"SAHs: Unraveling Safety Attention Heads in Large Vision-Language Models","authors":"Ziwei Zheng;Lijun He;Junyao Zhao;Le Yang;Fan Li","doi":"10.1109/TCSVT.2026.3710895","DOIUrl":"https://doi.org/10.1109/TCSVT.2026.3710895","url":null,"abstract":"With the integration of an additional modality, large vision-language models (LVLMs) exhibit greater vulnerability to safety risks (e.g., jailbreaking) compared to their language-only predecessors. Although recent studies have devoted considerable effort to the post-hoc alignment of LVLMs, the inner safety mechanisms remain largely unexplored. In this paper, we discover that activations of LVLM attention heads can pose significant linear separability of malicious and benign prompts. We hypothesize that the model’s inherent safety perception is governed by sparse attention heads, which we term “safety heads.” Further analysis reveals that these heads act as specialized shields against malicious prompts; ablating them leads to higher attack success rates, while the model’s utility remains unaffected, which further confirms our hypothesis. Beyond empirical evidence, we also conduct a semantic analysis showing that safety heads are capable of isolating and amplifying unsafe semantics in both image and text inputs, offering interpretable insights into their protective role. By locating these safety heads and concatenating their activations, we construct a straightforward but powerful malicious prompt detector that integrates seamlessly into the generation process with minimal extra inference overhead. Despite its simple structure of a logistic regression model, the detector surprisingly exhibits strong zero-shot generalization capabilities. Experiments across various prompt-based attacks in both image and video domains confirm the effectiveness of leveraging safety heads to protect LVLMs.","PeriodicalId":13082,"journal":{"name":"IEEE Transactions on Circuits and Systems for Video Technology","volume":"36 10","pages":"15317-15331"},"PeriodicalIF":10.8,"publicationDate":"2026-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149097906","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Blind Quality Enhancement of Compressed Video via Fine-Grained Degradation-Guided Sequential Inference 基于细粒度退化引导序列推理的压缩视频盲质量增强

IF 10.8 1区工程技术

IEEE Transactions on Circuits and Systems for Video Technology Pub Date : 2026-10-01 Epub Date: 2026-06-25 DOI: 10.1109/TCSVT.2026.3707192

Li Yu;Yingbo Zhao;Shiyu Wu;Siyue Yu;Moncef Gabbouj;Qingshan Liu

{"title":"Blind Quality Enhancement of Compressed Video via Fine-Grained Degradation-Guided Sequential Inference","authors":"Li Yu;Yingbo Zhao;Shiyu Wu;Siyue Yu;Moncef Gabbouj;Qingshan Liu","doi":"10.1109/TCSVT.2026.3707192","DOIUrl":"https://doi.org/10.1109/TCSVT.2026.3707192","url":null,"abstract":"Existing studies on quality enhancement for compressed video (QECV) predominantly rely on known quantization parameters (QPs), training separate enhancement models for each QP setting, which are referred to as non-blind methods. However, in practical scenarios such as transcoding and transmission, QPs may be partially or entirely unavailable, which limits the applicability of these methods and motivates the development of blind QECV techniques. Existing blind methods typically generate degradation vectors using classification models trained with cross-entropy loss, and employ them as channel attention to guide artifact reduction. Nevertheless, such degradation representations mainly capture global compression information and lack fine-grained spatial cues, making them less effective in handling spatially varying artifact patterns. To address this issue, we propose a pre-trained degradation representation learning module that decouples and extracts high-dimensional, multi-scale degradation representations from compressed video content, providing fine-grained guidance for artifact reduction. Furthermore, most existing blind and non-blind methods adopt a uniform inference architecture for all compression levels, ignoring the distinct computational demands of different QPs. To overcome this limitation, we introduce a sequential inference strategy that adaptively adjusts the number of artifact reduction stages according to the estimated compression level. Extensive experiments show that the proposed method significantly improves enhancement performance. In particular, at QP = 22, it raises PSNR improvement from 0.31 dB to 0.65 dB over the previous state-of-the-art blind method. Meanwhile, with the proposed sequential inference strategy, the average inference time at QP = 22 is reduced by 50% compared with that at QP = 42.","PeriodicalId":13082,"journal":{"name":"IEEE Transactions on Circuits and Systems for Video Technology","volume":"36 10","pages":"15421-15432"},"PeriodicalIF":10.8,"publicationDate":"2026-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11579375","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149097990","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Progressive Secret Image Sharing Using Multiplicative Primitive Irreducible Polynomials 基于乘法原始不可约多项式的渐进秘密图像共享

IF 10.8 1区工程技术

IEEE Transactions on Circuits and Systems for Video Technology Pub Date : 2026-10-01 Epub Date: 2026-06-17 DOI: 10.1109/TCSVT.2026.3704601

Xiaotian Wu;Xinyue Wang;Bing Chen;Zhihua Xia;Ching-Nung Yang;WeiQi Yan

{"title":"Progressive Secret Image Sharing Using Multiplicative Primitive Irreducible Polynomials","authors":"Xiaotian Wu;Xinyue Wang;Bing Chen;Zhihua Xia;Ching-Nung Yang;WeiQi Yan","doi":"10.1109/TCSVT.2026.3704601","DOIUrl":"https://doi.org/10.1109/TCSVT.2026.3704601","url":null,"abstract":"Progressive secret image sharing (PSIS) encodes a secret into <inline-formula> <tex-math>$n$ </tex-math></inline-formula> shadows, enabling the secret to be gradually revealed as more shared data becomes available. However, current techniques suffer from non-uniform progressive recovery and inferior visual quality of low-level decoded images. To address these issues, we combine multiplicative primitive irreducible polynomials with polynomial ring Chinese Remainder Theorem (CRT) to establish a PSIS. A method with 2 progressive levels is presented, where two polynomials <inline-formula> <tex-math>$f_{1}(alpha)$ </tex-math></inline-formula> and <inline-formula> <tex-math>$f_{2}(alpha)$ </tex-math></inline-formula> determine a main polynomial <inline-formula> <tex-math>$F(alpha) = f_{1}(alpha) times f_{2}(alpha)$ </tex-math></inline-formula>. A <inline-formula> <tex-math>$(k-1)$ </tex-math></inline-formula>-degree polynomial over <inline-formula> <tex-math>$(bmod ; F(alpha))$ </tex-math></inline-formula> is constructed for shadow generation. The secret can be losslessly recovered by Lagrange interpolation over <inline-formula> <tex-math>$(bmod ; F(alpha))$ </tex-math></inline-formula>, or gradually recovered through modified shadows using Lagrange interpolation and polynomial ring CRT. The scheme is further extended to <inline-formula> <tex-math>$t$ </tex-math></inline-formula> progressive levels, and a bit rearrangement technique is devised to enhance progressive recovery performance. Experiments and comparisons illustrate the effectiveness of the proposed technique, including uniform progressive recovery and superior low-level image quality.","PeriodicalId":13082,"journal":{"name":"IEEE Transactions on Circuits and Systems for Video Technology","volume":"36 10","pages":"15832-15847"},"PeriodicalIF":10.8,"publicationDate":"2026-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149097053","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

A Glimpse to Compress: Dynamic Visual Token Pruning for Large Vision-Language Models 压缩的一瞥：大型视觉语言模型的动态视觉标记修剪

IF 10.8 1区工程技术

IEEE Transactions on Circuits and Systems for Video Technology Pub Date : 2026-10-01 Epub Date: 2026-06-10 DOI: 10.1109/TCSVT.2026.3702147

Quan-Sheng Zeng;Yunheng Li;Qilong Wang;Peng-Tao Jiang;Zuxuan Wu;Ming-Ming Cheng;Qibin Hou

{"title":"A Glimpse to Compress: Dynamic Visual Token Pruning for Large Vision-Language Models","authors":"Quan-Sheng Zeng;Yunheng Li;Qilong Wang;Peng-Tao Jiang;Zuxuan Wu;Ming-Ming Cheng;Qibin Hou","doi":"10.1109/TCSVT.2026.3702147","DOIUrl":"https://doi.org/10.1109/TCSVT.2026.3702147","url":null,"abstract":"Visual token compression is critical for Large Vision-Language Models (LVLMs) to efficiently process high-resolution inputs. Existing methods that typically adopt fixed compression ratios cannot adapt to scenes of varying complexity, often causing imprecise pruning that discards informative visual tokens and results in degraded model performance. To address this issue, we introduce a dynamic pruning framework, GlimpsePrune, inspired by human cognition. It takes a data-driven “glimpse” and prunes irrelevant visual tokens in a single forward pass before answer generation. This approach prunes 92.4% of visual tokens while on average nearly preserving the original performance on free-form VQA tasks, causing end-to-end latency and peak memory to reduce to 66% and 67% of the baseline, respectively. By integrating our pruning mechanism into the reinforcement learning fine-tuning stage, the time and memory costs of fine-tuning are also reduced with negligible performance loss. Code is available at <uri>https://github.com/HVision-NKU/GlimpsePrune</uri>","PeriodicalId":13082,"journal":{"name":"IEEE Transactions on Circuits and Systems for Video Technology","volume":"36 10","pages":"15539-15552"},"PeriodicalIF":10.8,"publicationDate":"2026-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149097330","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

FaithViP: A Benchmark for Hallucinations With Synthetic Visual Prompts in Large Multi-Modal Models FaithViP：在大型多模态模型中使用合成视觉提示的幻觉基准

IF 10.8 1区工程技术

IEEE Transactions on Circuits and Systems for Video Technology Pub Date : 2026-10-01 Epub Date: 2026-07-01 DOI: 10.1109/TCSVT.2026.3708848

Yifei Gao;Ning Xu;Hongshuo Tian;Sha Wei;An-An Liu

{"title":"FaithViP: A Benchmark for Hallucinations With Synthetic Visual Prompts in Large Multi-Modal Models","authors":"Yifei Gao;Ning Xu;Hongshuo Tian;Sha Wei;An-An Liu","doi":"10.1109/TCSVT.2026.3708848","DOIUrl":"https://doi.org/10.1109/TCSVT.2026.3708848","url":null,"abstract":"Visual prompts serve as a representative type of synthetic objects, showing promise for guiding large multi-modal models (LMMs) to achieve precise region-specific understanding in applications like medical imaging and autonomous driving. However, the free synthetic form and pixel-level granularity of visual prompts make them a significant hallucination source. Despite existing benchmarks for general object hallucination, there is a critical need for a dedicated benchmark to evaluate the faithfulness of LMMs when faced with diverse synthetic prompts. In this paper, we introduce FaithViP, a comprehensive synthetic-visual-prompt benchmark to assess the faithfulness for human-interactive reasoning in LMMs. Drawing insights from human cognition, we design diverse hallucination scenarios, including missing or ambiguous visual prompts and object references, attribute conflicts between instruction and marked region. FaithViP comprises 308 meticulously crafted image-question pairs that induce visual-prompt hallucinations. We evaluate a wide range of LMM baselines, including general LMMs and those fine-tuned on visual-prompt data. Our results show that even SOTA LMMs like GPT-4o, Qwen2.5-VL and ViP-LLaVA series struggle to achieve optimal performance, showing the challenge of our FaithViP. Furthermore, we analyze the underlying causes of visual-prompt hallucinations, uncovering the limitations of existing open-source LMMs. Our work provides key insights into enhancing visual-textual interaction capabilities for LMMs, advancing more robust real-world applications. Our code and data are available at <uri>https://github.com/YFCodeDream/FaithViP</uri>","PeriodicalId":13082,"journal":{"name":"IEEE Transactions on Circuits and Systems for Video Technology","volume":"36 10","pages":"15639-15651"},"PeriodicalIF":10.8,"publicationDate":"2026-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149097398","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

FedDD: Combating Federated Noisy Label Learning With Intra- and Inter-Distributional Difference 利用分布内和分布间差异对抗联邦噪声标签学习

IF 10.8 1区工程技术

IEEE Transactions on Circuits and Systems for Video Technology Pub Date : 2026-10-01 Epub Date: 2026-06-05 DOI: 10.1109/TCSVT.2026.3700385

Tong Jin;Chen-Chen Zong;Sheng-Jun Huang

{"title":"FedDD: Combating Federated Noisy Label Learning With Intra- and Inter-Distributional Difference","authors":"Tong Jin;Chen-Chen Zong;Sheng-Jun Huang","doi":"10.1109/TCSVT.2026.3700385","DOIUrl":"https://doi.org/10.1109/TCSVT.2026.3700385","url":null,"abstract":"Federated learning (FL) enables multiple parties to collaboratively train a global machine learning model without sharing local data. While FL has made awe-inspiring advances, most existing studies assume fully clean data labels on clients, which rarely holds in real-world scenarios where label noise is prevalent and detrimental to FL performance. Prior approaches typically rely on the small-loss criterion to identify noisy instances, which has been proven to struggle in highly heterogeneous settings. To overcome this, we propose a robust framework called <bold>FedDD</b> that tackles label noise in FL by leveraging intra- and inter-distributional differences. Instead, FedDD employs an intra-difference measure, which utilizes the intra-logit variance to preliminarily identify the clean subset. Subsequently, it refines the data by introducing an inter-distributional difference metric based on prediction discrepancies between local and global models. To further utilize identified noisy examples, a novel federated semi-supervised learning paradigm is proposed that amalgamates knowledge from the global and local models to guide pseudo-labeling. Extensive experiments on benchmark datasets under both independent and identically distributed (IID) and non-IID settings demonstrate the superiority of FedDD over state-of-the-art methods.","PeriodicalId":13082,"journal":{"name":"IEEE Transactions on Circuits and Systems for Video Technology","volume":"36 10","pages":"14897-14911"},"PeriodicalIF":10.8,"publicationDate":"2026-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149097547","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

DTTrack: Metric-Guided Dynamic Template Tracking Network DTTrack：度量导向的动态模板跟踪网络

IF 10.8 1区工程技术

IEEE Transactions on Circuits and Systems for Video Technology Pub Date : 2026-10-01 Epub Date: 2026-07-03 DOI: 10.1109/TCSVT.2026.3709830

Yijun Liang;Ziyi Chen;Yewang Chen;Yun Chen;Dilong Li;Weiquan Liu;Cheng Wang;Jonathan Li

{"title":"DTTrack: Metric-Guided Dynamic Template Tracking Network","authors":"Yijun Liang;Ziyi Chen;Yewang Chen;Yun Chen;Dilong Li;Weiquan Liu;Cheng Wang;Jonathan Li","doi":"10.1109/TCSVT.2026.3709830","DOIUrl":"https://doi.org/10.1109/TCSVT.2026.3709830","url":null,"abstract":"While unified weight-sharing architectures have shown strong versatility across RGB-T, RGB-D, and RGB-E tracking, accommodating heterogeneous data often weakens fine-grained target discriminability and inevitably leads to feature dilution in long-term tracking. Conventional update strategies also struggle to maintain reliable memory under frequent shifts in modality dominance and heterogeneous noise, thereby limiting tracking performance. To address these issues, this paper proposes the Metric-Guided Dynamic Template Tracking Network (DTTrack). By constructing an independent, modality-agnostic metric space, DTTrack enables effective long-term memory modeling in a single-stream framework without modifying the backbone. Specifically, a Metric-Guided Synergistic Decision Head (MSDH) is introduced to decouple localization and discriminative features explicitly. Additionally, a Sequence-Aware Adaptive Sampling (SeAS) strategy is developed to encourage identity-invariant learning across temporal scales. Furthermore, a quality- and diversity-aware dynamic update mechanism is designed in the learned metric space to select high-confidence templates and suppress heterogeneous noise. Extensive experiments on five benchmark datasets across RGB-T, RGB-D, and RGB-E tracking demonstrate that DTTrack achieves state-of-the-art or highly competitive performance.","PeriodicalId":13082,"journal":{"name":"IEEE Transactions on Circuits and Systems for Video Technology","volume":"36 10","pages":"16065-16080"},"PeriodicalIF":10.8,"publicationDate":"2026-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149097634","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Robust Dataset Pruning via Joint Noise-Aware Discrimination and Re-Labeling 基于联合噪声感知识别和重标记的鲁棒数据集修剪

IF 10.8 1区工程技术

IEEE Transactions on Circuits and Systems for Video Technology Pub Date : 2026-10-01 Epub Date: 2026-06-29 DOI: 10.1109/TCSVT.2026.3708310

Weiwei Xiao;Yi Ma;Shaocong Wu;Qiben Shan;Yaowei Wang;Jingyong Su

{"title":"Robust Dataset Pruning via Joint Noise-Aware Discrimination and Re-Labeling","authors":"Weiwei Xiao;Yi Ma;Shaocong Wu;Qiben Shan;Yaowei Wang;Jingyong Su","doi":"10.1109/TCSVT.2026.3708310","DOIUrl":"https://doi.org/10.1109/TCSVT.2026.3708310","url":null,"abstract":"Deep neural networks have achieved success, but require enormous computational costs and storage space on large-scale datasets. Dataset pruning is an effective method to construct compact subsets from large-scale datasets to reduce computational costs while preserving model performance. Existing dataset pruning methods predominantly focus on clean benchmark datasets, overlooking the critical challenges of real-world scenarios with prevalent noisy label samples. In this paper, we find that previous methods have the limitation of being unable to distinguish noisy label samples from hard samples, resulting in overfitting to noisy label samples. To address this, we propose Robust Pruning (RoP), a two-stage dataset pruning framework that consists of noisy label discrimination and re-labeling. Specifically, RoP employs the divergence in label distributions between given labeled samples and their neighboring predictions as an importance metric to identify noisy samples. Compared to other neighborhood-based methods, RoP corrects uncertain neighboring samples through feature and label propagation. Subsequently, RoP applies re-labeling techniques to the selected subset to prevent overfitting of noisy labels. Finally, RoP evaluates easy and hard samples through empirical analysis of re-labeling accuracy, guaranteeing the coverage of hard and easy samples in the selected subset. Extensive experiments demonstrate the effectiveness of the proposed method, not only on real-world benchmarks but also on synthetic datasets, highlighting its suitability for noisy label scenarios. The code is available at <uri>https://github.com/successhaha/RoP</uri>","PeriodicalId":13082,"journal":{"name":"IEEE Transactions on Circuits and Systems for Video Technology","volume":"36 10","pages":"15206-15219"},"PeriodicalIF":10.8,"publicationDate":"2026-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149097753","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Hierarchical Sparse Graph for Multi-Dimensional Assignment in Event-Based Tracking 基于事件跟踪的多维分配层次稀疏图

IF 10.8 1区工程技术

IEEE Transactions on Circuits and Systems for Video Technology Pub Date : 2026-10-01 Epub Date: 2026-06-17 DOI: 10.1109/TCSVT.2026.3704671

Muxi Zha;Yang Shang;Wenlong Zhang;Qifeng Yu;Banglei Guan

{"title":"Hierarchical Sparse Graph for Multi-Dimensional Assignment in Event-Based Tracking","authors":"Muxi Zha;Yang Shang;Wenlong Zhang;Qifeng Yu;Banglei Guan","doi":"10.1109/TCSVT.2026.3704671","DOIUrl":"https://doi.org/10.1109/TCSVT.2026.3704671","url":null,"abstract":"Existing event-based tracking methods typically operate in an online manner, associating tracklets across adjacent batches. However, this limited temporal window often leads to failures in challenging scenarios, such as long-term occlusions or missed detections, where broader temporal context should be incorporated for high-precision tracking. To achieve robust and accurate tracking, this paper concentrates on the event-based multi-dimensional assignment that associates tracklets across multiple batches simultaneously. Unlike image-based detectors, event-based detectors can generate multiple correct yet spatially distinct tracklets for a single object per batch. This violates the one-to-one rule in classic data association, making event-based association an underdetermined many-to-many matching problem. To the best of our knowledge, this is the first work to tackle this NP-hard problem. To address its underdetermined nature, we formulate the assignment problem via fuzzy logic, where pairwise tracklet correlations are quantified through multiple long-range kinematic metrics. This fuzzy design eliminates the need for hard tracking thresholds and enables multiple heterogeneous metrics to be integrated into a unified correlation. By treating tracklets as nodes and their fuzzy correlations as edges, we construct a Hierarchical Sparse Graph (HSG). Building upon the HSG, we pose assignment as a graph partition task, and employ a heuristic solution that measures vertex affinity for partitioning adaptively. The entire pipeline is training-free. Comprehensive experiments on two challenging event datasets, eTram and Ev-UAV, show that our method outperforms existing event-based and most image-based association models. Particularly, it achieves over <inline-formula> <tex-math>$2times $ </tex-math></inline-formula> the performance of OC-SORT across all metrics in the many-to-many association scenario.","PeriodicalId":13082,"journal":{"name":"IEEE Transactions on Circuits and Systems for Video Technology","volume":"36 10","pages":"14755-14770"},"PeriodicalIF":10.8,"publicationDate":"2026-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149096974","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0