检索结果-内蒙古大学图书馆

Exploring the Better Correlation for Few-Shot video object segmentation

IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR video TECHNOLOGY 2025年第3期35卷 2133-2146页

作者： Luo, Naisong Wang, Yuan Sun, Rui Xiong, Guoxin Zhang, Tianzhu Wu, Feng Univ Sci & Technol China Sch Informat Sci Hefei 230027 Peoples R China Deep Space Explorat Lab Hefei 230088 Peoples R China

Few-shot video object segmentation (FSVOS) aims to achieve accurate segmentation of novel objects in given video sequences, where the target objects are specified by limited annotated images as support. Most previous top-performing methods adopt the support-query semantic correlation learning paradigm or the intra-query temporal correlation learning paradigm. Nevertheless, they either fail to model temporal consistency across frames, resulting in inconsecutive segmentation, or lose diverse support object information, leading to incomplete segmentation. Therefore, we argue that it is more desirable to achieve both correlations in a collaborative manner. In this work, we delve into the issues present in the combination of few-shot image segmentation methods and video object segmentation methods and propose a dedicated Collaborative Correlation Network (CoCoNet) to address these problems, including a pixel correlation calibration module and a temporal correlation mining module. The proposed CoCoNet enjoys several merits. First, the pixel correlation calibration module aims to mitigate the noise issue in support-query correlation by integrating the affinity learning strategy and the prototype learning strategy. Specifically, we employ Optimal Transport to enrich pixel correlation with contextual information, thereby reducing intra-class differences between support and query. Second, the temporal correlation mining module is responsible for alleviating the issue of uncertainty in the initial frame and establishing reliable guidance for subsequent frames of the query video. With the collaboration of these two modules, our CoCoNet can effectively establish support-query and temporal correlation simultaneously and achieve accurate FSVOS. Extensive experimental results on two challenging benchmarks demonstrate that our method performs favorably against state-of-the-art FSVOS methods.

关键词： Few-shot video object segmentation video object segmentation few-shot learning Few-shot video object segmentation video object segmentation few-shot learning

来源：评论

学校读者我要写书评

暂无评论

Temporal Transductive Inference for Few-Shot video object segmentation

引用

INTERNATIONAL JOURNAL OF COMPUTER VISION 2025年 1-18页

作者： Siam, Mennatullah Univ British Columbia Comp Sci Vancouver BC Canada

Few-shot video object segmentation (FS-VOS) aims at segmenting video frames using a few labelled examples of classes not seen during initial training. In this paper, we present a simple but effective temporal transductive inference (TTI) approach that leverages temporal consistency in the unlabelled video frames during few-shot inference without episodic training. Key to our approach is the use of a video-level temporal constraint that augments frame-level constraints. The objective of the video-level constraint is to learn consistent linear classifiers for novel classes across the image sequence. It acts as a spatiotemporal regularizer during the transductive inference to increase temporal coherence and reduce overfitting on the few-shot support set. Empirically, our approach outperforms state-of-the-art meta-learning approaches in terms of mean intersection over union on YouTube-VIS by 2.5%. In addition, we introduce an improved benchmark dataset that is exhaustively labelled (i.e., all object occurrences are labelled, unlike the currently available). Our empirical results and temporal consistency analysis confirm the added benefits of the proposed spatiotemporal regularizer to improve temporal coherence. Our code and benchmark dataset is publicly available at, https://***/MSiam/tti_fsvos/.

关键词： Few-shot learning Transductive inference video object segmentation

来源：评论

学校读者我要写书评

暂无评论

MoBox: Enhancing video object segmentation With Motion-Augmented Box Supervision

引用

IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR video TECHNOLOGY 2025年第1期35卷 405-417页

作者： Li, Xiaomin Wang, Qinghe Li, Dezhuang Ge, Mengmeng Jia, Xu He, You Lu, Huchuan Dalian Univ Technol Sch Artificial Intelligence Dalian 116024 Peoples R China Dalian Univ Technol Sch Informat & Commun Engn Dalian 116024 Peoples R China Tsinghua Univ Dept Comp Sci & Technol Beijing 100084 Peoples R China Tsinghua Univ Dept Elect Engn Beijing 100084 Peoples R China

We propose MoBox, a low-cost solution for semi-supervised video object segmentation that requires only bounding boxes as manual annotations for training. Built upon a mature semi-supervised video object segmentation network, we redesign the training losses and employ a more stringent training strategy. Specifically, we introduce a well-designed constraint term that enhances traditional spatial projection by simultaneously leveraging the projections of both the ground-truth box and the predicted mask across two axes, rather than evaluating discrepancies along the x-axis and y-axis independently. To harness the intrinsic properties of videos, considering the underlying correspondence between motion represented by optical flow and the original image, we incorporate motion coherence information into the color consistency loss as supplementary information and propose a motion discrepancy loss to obtain accurate boundaries. Additionally, to mitigate the ambiguity of weak supervision, we further introduce the pseudo strict constraint during training, which significantly improves model performance. Our approach yields competitive scores on popular benchmarks, achieving a J & F score of 78.6 on the DAVIS 2017 validation set and an Overall score of 78.0 on the YouTube-VOS 2018 validation set. These results highlight the efficacy of MoBox, demonstrating that the semi-supervised video object segmentation model can be effectively trained using only motion-augmented box supervision and intrinsic information of videos.

关键词： Annotations Training Optical flow Task analysis object segmentation Image color analysis Circuits and systems video object segmentation box annotations motion constraint optical flow

来源：评论

学校读者我要写书评

暂无评论

Self-supervised video object segmentation via pseudo label rectification

引用

PATTERN RECOGNITION 2025年 163卷

作者： Guo, Pinxue Zhang, Wei Li, Xiaoqiang Fan, Jianping Zhang, Wenqiang Fudan Univ Acad Engn & Technol Shanghai Peoples R China Fudan Univ Sch Comp Sci Shanghai Peoples R China Shanghai Key Lab Intelligent Informat Proc Shanghai Peoples R China Shanghai Univ Sch Comp Engn & Sci Shanghai Peoples R China Lenovo Res AI Lab Beijing Peoples R China

In this paper we propose a novel self-supervised framework for video object segmentation (VOS) which consists of siamese encoders and bi-decoders. Siamese encoders extract multi-level features and generate pseudo labels for each pixel by cross attention in visual-semantic space. Such siamese encoders are learned via the colorization task without any labeled video data. Bi-decoders take in features from different layers of the encoder and output refined segmentation masks. Such bi-decoders are trained by the pseudo labels, and in turn pseudo labels are rectified via bi-decoders mutual learning. The variation of the bi-decoders' outputs is minimized such that the gap between pseudo labels and the ground-truth is reduced. Experimental results on the challenging datasets DAVIS-2017 and YouTube-VOS demonstrate the effectiveness of our proposed approach.

关键词： video object segmentation Self-supervised learning Pseudo label learning

来源：评论

学校读者我要写书评

暂无评论

Vanishing mask refinement in semi-supervised video object segmentation

引用

APPLIED SOFT COMPUTING 2025年 172卷

作者： Pita, Javier Llerena, Juan P. Patricio, Miguel A. Berlanga, Antonio Usero, Luis Grp MasMovil MasMovil Team Ave Bruselas 38 Madrid 28108 Spain Univ Alcala Cognit Sci Res Grp Ctra Madrid Barcelona km Campus UnivCtra Madrid Barcelona km33600 Madrid 28805 Spain Univ Carlos III Madrid Comp Sci & Engn Dept Appl Artificial Intelligence Grp Avda Gregorio Peces-Barba & Martinez22Colmenarej Madrid 28270 Spain

This paper presents video object segmentation Enhanced with Segment Anything Model (VOS-E-SAM), a multistage architecture for Semi-supervised video object segmentation (SVOS) using the foundational Segment Anything Model (SAM) architecture, aimed at addressing the challenges of mask degradation over time in long video sequences. Our architectural approach enhances the object masks produced by the XMem model by incorporating SAM. This integration uses various input combinations and low-level computer vision techniques to generate point prompts, in order to improve mask continuity and accuracy throughout the entire video cycle. The main challenge addressed is the fading or vanishing of object masks in long video sequences due to problems such as changes in object appearance, occlusions, camera movements, and approach changes. Both the baseline architecture and the newer high-quality version are tested, addressing the primary challenge of object mask fading or vanishing in long video sequences due to changes in object appearance, occlusions, camera movements, and variations in approach. Through rigorous experimentation with different prompt configurations, we identified an outstanding configuration of SAM inputs to improve mask refinement. Evaluations on benchmark long video datasets, such as LongDataset and LVOS, show that our approach significantly improves mask quality in single-object extended video sequences proven by percentage increments on jaccard index (J ) and contour accuracy (F) based metrics (mean, recall and decay). Our results show remarkable improvements in mask persistence and accuracy, which sets a new standard for the integration of foundational models in video segmentation and lays the foundation for future research in this field. ***-E-SAM

关键词： video object segmentation Foundation model object segmentation Long-term videos Deep learning

来源：评论

学校读者我要写书评

暂无评论

Unsupervised video object segmentation with mask transformer: boosting accuracy and efficiency through feature fusion

引用

VISUAL COMPUTER 2025年 1-12页

作者： Qu, Daikun Zhao, Hongwei Zhou, Mingzhu Jilin Univ Coll Comp Sci & Technol Changchun 130012 Peoples R China Jilin Univ Key Lab Symbol Computat & Knowledge Engn Minist Educ Changchun 130012 Peoples R China Jilin Univ China Japan Union Hosp Changchun 130033 Peoples R China

This paper introduces a novel Mask Transformer framework for unsupervised video object segmentation, focusing on enhancing segmentation accuracy and efficiency. Our approach leverages the power of transformers to capture long-range dependencies and spatial context, enabling precise segmentation of objects in videos without requiring manual annotations during training. By fusing appearance and motion features through a unique combination of pixel-level feature representation and object-level query processing, our model achieves state-of-the-art performance on benchmark datasets while maintaining a substantial training speed improvement. The proposed framework offers a significant step forward in the field, demonstrating the effectiveness of mask classification for UVOS tasks. We open-source our code and datasets to facilitate reproducibility and encourage further research in this area at https://***/Qudk23/MaskVOS?tab=readme-ov-file.

关键词： Computer vison video object segmentation Transformer Mask classification

来源：评论

学校读者我要写书评

暂无评论

SKVOS: Sketch-Based video object segmentation with a Large-Scale Benchmark

引用

APPLIED SCIENCES-BASEL 2025年第4期15卷 1751-1751页

作者： Yang, Ruolin Li, Da Hu, Conghui Zhang, Honggang Beijing Univ Posts & Telecommun Sch Artificial Intelligence Beijing 100876 Peoples R China Univ Surrey Ctr Vis Speech & Signal Proc SketchX Surrey GU2 7XH England Natl Univ Singapore Dept Comp Sci Singapore 119077 Singapore

In this paper, we propose sketch-based video object segmentation (SKVOS), a novel task that segments objects consistently across video frames using human-drawn sketches as queries. Traditional reference-based methods, such as photo masks and language descriptions, are commonly used for segmentation. Photo masks provide high precision but are labor intensive, limiting scalability. While language descriptions are easy to provide, they often lack the specificity needed to distinguish visually similar objects within a frame. Despite their simplicity, sketches capture rich, fine-grained details of target objects and can be rapidly created, even by non-experts, making them an attractive alternative for segmentation tasks. We introduce a new approach that utilizes sketches as efficient and informative references for video object segmentation. To evaluate sketch-guided segmentation, we introduce a new benchmark consisting of three datasets: Sketch-DAVIS16, Sketch-DAVIS17, and Sketch-YouTube-VOS. Building on a memory-based framework for semi-supervised video object segmentation, we explore effective strategies for integrating sketch-based references. To ensure robust spatiotemporal coherence, we introduce two key innovations: the Temporal Relation Module and Sketch-Anchored Contrastive Learning. These modules enhance the model's ability to maintain consistency both across time and across different object instances. Our method is evaluated on the Sketch-VOS benchmark, demonstrating superior performance with overall improvements of 1.9%, 3.3%, and 2.0% over state-of-the-art methods on the Sketch-YouTube-VOS, Sketch-DAVIS 2016, and Sketch-DAVIS 2017 validation sets, respectively. Additionally, on the YouTube-VOS validation set, our method outperforms the leading language-based VOS approach by 10.1%.

关键词： sketches video object segmentation sketch-based datasets

来源：评论

学校读者我要写书评

暂无评论

MAHC: motion-appearance video object segmentation via hierarchical attention and multi-level clustering

引用

JOURNAL OF SUPERCOMPUTING 2025年第5期81卷 1-30页

作者： Cao, Honghui Yang, Yang Cao, Lvchen Liu, Chenyang Hou, Yandong Wang, Jun Henan Univ Sch Artificial Intelligence 379 Mingli Rd North Zhengzhou 450046 Henan Peoples R China Henan Univ Technol LuoHe Vocat Technol Coll 100 Lianhua StHigh Tech Dev Zone Zhengzhou 450001 Henan Peoples R China Henan JinShu Intelligence Technol Co Ltd Zhengzhou Henan Peoples R China

Unsupervised video object segmentation (UVOS) aims to automatically segment the most salient and semantically meaningful objects in a video without relying on manual annotations. Existing methods often focus on direct feature fusion without fully exploiting the inherent advantages of individual features, leading to limited performance when handling scenes with diverse motion patterns or similar foreground-background appearances. To address these challenges, we propose MAHC (motion-appearance hierarchical clustering), a novel unsupervised video object segmentation framework that effectively integrates motion and appearance cues through hierarchical feature learning and progressive clustering refinement. Our framework employs a hierarchical interleaved attention mechanism within an autoencoder structure to enhance motion feature representation, and utilizes a multi-level clustering strategy that progressively integrates different clustering techniques to achieve comprehensive segmentation from global patterns to fine-grained details. Additionally, we improve foreground-background discrimination by combining multi-view subspace analysis with motion intensity information from reconstructed optical flow. Extensive experiments on four challenging benchmark datasets (DAVIS16, FBMS, SegTrackv2, and YouTube-objects) demonstrate that MAHC significantly outperforms existing unsupervised methods.

关键词： video object segmentation Hierarchical interleaved attention Multi-level clustering refinement Motion-appearance fusion

来源：评论

学校读者我要写书评

暂无评论

Scribble-Supervised video object segmentation via Scribble Enhancement

引用

IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR video TECHNOLOGY 2025年第4期35卷 2999-3012页

作者： Gao, Xingyu Li, Zuolei Shi, Hailong Chen, Zhenyu Zhao, Peilin Chinese Acad Sci Inst Microelect Beijing 100029 Peoples R China State Grid Corp China Big Data Ctr Beijing 100031 Peoples R China China Elect Power Res Inst Beijing 100192 Peoples R China Tencent AI Lab Shenzhen 518000 Peoples R China

Current video object segmentation methods heavily rely on pixel-level mask annotations when training, which are expensive and time-consuming to acquire. To address this problem, some approaches try to train with sparse scribble annotations and take sparse target scribble as initial information for inference. However, due to the sparsity of scribble annotations, the performance is often limited, and the corresponding loss function needs to be designed. Inspired by the powerful ability of Segment Anything Model (SAM) to leverage prompt for segmentation, we argue that this problem can be alleviated by improving the quality of scribble. Therefore, we propose SEVOS, a framework for scribble-supervised video object segmentation, which contains a scribble enhancement algorithm and an semi-supervised video object segmentation network. Specifically, the scribble enhancement algorithm first samples corresponding positive sample points and negative sample points from target scribbles, and then feeds them into the SAM in turn, achieving high-quality scribble enhancement without human intervention. This algorithm augments the scribble-annotated video dataset, which is used for additional training of the model. Furthermore, we design a post-processing enhancement algorithm to further improve the prediction results. The obtained model outperforms state-of-the-art methods with a considerable performance gap, indicating the generalization and effectiveness of the proposed model.

关键词： Annotations Training object segmentation Prediction algorithms Circuits and systems Reliability Accuracy Costs Transformers Target tracking video object segmentation weakly supervised segment anything scribble enhancement

来源：评论

学校读者我要写书评

暂无评论

PRM: A Pixel-Region-Matching Approach for Fast video object segmentation 7th

PRM: A Pixel-Region-Matching Approach for Fast Video Object ...

引用

7th Chinese Conference on Pattern Recognition and Computer Vision

作者： Wang, Zhen Zhang, Wenbin Li, Jiahao Zhu, Wencheng Song, Danqing Hebei Univ Technol Tianjin Peoples R China Tsinghua Univ Beijing Peoples R China South China Univ Technol Guangzhou Peoples R China

ISBN: (纸本)9789819785049;9789819785056

We present a pixel-region-matching approach for fast semi-supervised video object segmentation. Unlike existing methods that exploit pixel-wise similarities between search and query inputs, our method simultaneously extracts pixel-level and region-level information. Specifically, our method proposes a pixel-wise similarity module and a region-based correlation module for video object segmentation. The region-based correlation module learns object prototypes via appearance features that pay more attention to the most relevant target pixels, while the pixel-wise similarity module obtains subtly discriminative information for fine-grained object segmentation. Finally, we integrate these features by using the proposed pixel-region fusion module and then feed them into decoder networks to predict foreground and background. Notably, our method is an end-to-end trainable framework without online fine-tuning and post-processing. We evaluate our method on three widely-used datasets including DAVIS16, DAVIS17, and YouTube-VOS. Experimental results clearly demonstrate the effectiveness of the proposed method compared with state-of-the-art methods.

关键词： video object segmentation pixel-wise similarity region-level matching pixel-region fusion

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：