109,407 Hits in 2.5 sec

3D video performance segmentation

Tony Tung, Takashi Matsuyama
2010 2010 IEEE International Conference on Image Processing  
Our method allows us to perform robust body part segmentation and tracking in 3D cinema sequences.  ...  We present a novel approach that achieves segmentation of subject body parts in 3D videos. 3D video consists in a freeviewpoint video of real-world subjects in motion immersed in a virtual world.  ...  Fig. 4 . 4 Motion graph. 3D video sequences are clustered in topology classes. Fig. 5 . 5 3D video performance segmentation.  ... 
doi:10.1109/icip.2010.5652541 dblp:conf/icip/TungM10 fatcat:xft332w3grgefmwcqq3xnttclu

Making a Case for 3D Convolutions for Object Segmentation in Videos [article]

Sabarinath Mahadevan, Ali Athar, Aljoša Ošep, Sebastian Hennen, Laura Leal-Taixé, Bastian Leibe
2020 arXiv   pre-print
In this work, we show that 3D CNNs can be effectively applied to dense video prediction tasks such as salient object segmentation.  ...  of videos compared to their 2D convolutional counterparts and lag behind the aforementioned networks in terms of performance.  ...  [15] was the first attempt to apply a fully 3D CNN for video object segmentation. While promising, it is out-performed by state-of-the-art methods that use 2D CNNs [47, 57] .  ... 
arXiv:2008.11516v1 fatcat:qhgfa4zohzgc7oyocu2gnwneqa

Identifying Same Persons from Temporally Synchronized Videos Taken by Multiple Wearable Cameras

Kang Zheng, Hao Guo, Xiaochuan Fan, Hongkai Yu, Song Wang
2016 2016 IEEE Conference on Computer Vision and Pattern Recognition Workshops (CVPRW)  
Specifically, we take the tracked persons from different videos to be the same person if their 3D poses are the same, given that these videos are synchronized.  ...  We adapt an existing algorithm to estimate the tracked person's 3D poses in each 2D video using motionbased features.  ...  For each video, we temporally divide it into segments. We perform nCTE for each segment to obtain top-K candidates of 3D pose sequences.  ... 
doi:10.1109/cvprw.2016.106 dblp:conf/cvpr/ZhengGFYW16 fatcat:ozhr3h75gvgorb7rwrrxwpqmae

Coronary Artery Segmentation in Angiographic Videos Using A 3D-2D CE-Net [article]

Lu Wang, Dong-xue Liang, Xiao-lei Yin, Jing Qiu, Zhi-yun Yang, Jun-hui Xing, Jian-zeng Dong, Zhao-yuan Ma
2020 arXiv   pre-print
This framework combines a 3D convolutional layer to extract spatial--temporal information from a video sequence and a 2D CE--Net to accomplish the segmentation task of an image sequence.  ...  From the results of segmentation and extraction, we can get good segmentation results despite the poor quality of coronary angiography video sequences.  ...  Table 1 . 1 Comparison of the performances of 2D CE-Net, 3D-2D CE-Net with N = 1 and N = 2 on the test set of SetV. . 2.Six examples of blood vessel segmentation from coronary artery angiographic videos  ... 
arXiv:2003.11851v3 fatcat:oleewbwl5vcbxdzfhml626itza

Evaluation of super-voxel methods for early video processing

Chenliang Xu, J. J. Corso
2012 2012 IEEE Conference on Computer Vision and Pattern Recognition  
Supervoxel segmentation has strong potential to be incorporated into early video analysis as superpixel segmentation has in image analysis.  ...  Our findings have led us to conclusive evidence that the hierarchical graph-based and segmentation by weighted aggregation methods perform best and almost equally-well on nearly all the metrics and are  ...  Given a ground-truth segmentation into segments g 1 , g 2 , ..., g m , and a video segmentation into supervoxels s 1 , s 2 , ..., s n , here we propose a suite of the volumetric video-based 3D metrics.  ... 
doi:10.1109/cvpr.2012.6247802 dblp:conf/cvpr/XuC12 fatcat:c7w7d2hlazgfdpmv57eekghxzy

3D Human Video Retrieval: from Pose to Motion Matching [article]

Rim Slama, Hazem Wannous, Mohamed Daoudi
2013 Eurographics Workshop on 3D Object Retrieval, EG 3DOR  
Moreover, local 3D video retrieval is performed by dynamic time warping (DTW) algorithm in the feature space vectors.  ...  Experiments on both synthetic and real 3D human video sequences show that our approach provides an accurate shape similarity in video compared to the best state-of-the-art approaches.  ...  We extended this descriptor to 3D video retrieval, where a motion segmentation is performed on continuous sequence to split it into elementary action segments called clips.  ... 
doi:10.2312/3dor/3dor13/033-040 fatcat:umwjsg5q5ffllfb5somvcinx3y

Dynamic Hyperlinker: Innovative Solution for 3D Video Content Search and Retrieval

Mohammad Rafiq Swash, Amar Aggoun, Obaidullah Abdul Fatah, Bei Li
2016 Journal of Computer and Communications  
It advances 3D multimedia navigability and searchability by creating dynamic links for selectable and clickable objects in the video scene whilst the user consumes the 3D video clip.  ...  The experiment is conducted on full parallax holoscopic 3D videos "also known as integral images".  ...  Center-views and 3D depth map are generated from H3D videos that are used for segmenting 3D objects as well as creating bounding boxes for 3D objects.  ... 
doi:10.4236/jcc.2016.46002 fatcat:6kdooaeubje5xeiiastga7j36q

Spatiotemporal Multi-Task Network for Human Activity Understanding

Yao Liu, Jianqiang Huang, Chang Zhou, Deng Cai, Xian-Sheng Hua
2017 Proceedings of the on Thematic Workshops of ACM Multimedia 2017 - Thematic Workshops '17  
To tackle these problems, we propose a spatiotemporal, multi-task, 3D deep convolutional neural network to detect (including temporally localize and recognition) actions in untrimmed videos.  ...  Meanwhile, in order to better utilize the rich motion information in videos, we introduce a novel video representation, interlaced images, as an additional network input stream.  ...  3D-Pool3 3D-Conv4a 3D-Conv4b 3D-Pool4 3D-Conv5a 3D-Conv5b FC6 FC7 Fusion FC8 Loss Segment-1 Share Weights Share Weights Segment-2 Segment-3 Figure 2 : Overview of our fusion architecture.  ... 
doi:10.1145/3126686.3126705 dblp:conf/mm/LiuHZCH17 fatcat:vauwsk6ndbd7fhxhel3swqf4fq

Video-object segmentation and 3D-trajectory estimation for monocular video sequences

Feng Xu, Kin-Man Lam, Qionghai Dai
2011 Image and Vision Computing  
It can be employed for video-object segmentation, 2D-to-3D video conversion, video-object retrieval, etc.  ...  Experiments show that the proposed framework can achieve a good performance in terms of both object segmentation and 3D-trajectory estimation.  ...  For 2D-to-3D video conversion, as the 3D property of a scene changes with time and is captured by one camera, 3D reconstruction has to be performed for every frame.  ... 
doi:10.1016/j.imavis.2010.09.001 fatcat:ezcdsgofizgrxfdylswpe4ayre

Video object cut and paste

Yin Li, Jian Sun, Heung-Yeung Shum
2005 ACM Transactions on Graphics  
To achieve this, we first apply a new 3D graph cut based segmentation approach on the spatialtemporal video volume.  ...  Then, the initial segmentation result is refined locally.  ...  Many thanks to Yingqing Xu and Beijing Dancing Academy for help on video sources.  ... 
doi:10.1145/1073204.1073234 fatcat:xmweyehddjcnlmgzyuhhqlgn4i

Video object cut and paste

Yin Li, Jian Sun, Heung-Yeung Shum
2005 ACM SIGGRAPH 2005 Papers on - SIGGRAPH '05  
To achieve this, we first apply a new 3D graph cut based segmentation approach on the spatialtemporal video volume.  ...  Then, the initial segmentation result is refined locally.  ...  Many thanks to Yingqing Xu and Beijing Dancing Academy for help on video sources.  ... 
doi:10.1145/1186822.1073234 fatcat:7yqby4tzcjctxetw433lc4nity

Coronary artery segmentation in angiographic videos utilizing spatial-temporal information

Lu Wang, Dongxue Liang, Xiaolei Yin, Jing Qiu, Zhiyun Yang, Junhui Xing, Jianzeng Dong, Zhaoyuan Ma
2020 BMC Medical Imaging  
Methods This article proposes a novel coronary artery segmentation framework that combines a three–dimensional (3D) convolutional input layer and a two–dimensional (2D) convolutional network.  ...  Segmentation and extraction of blood vessels from coronary angiographic images or videos are very essential prerequisites for physicians to locate, assess and diagnose the plaques and stenosis in blood  ...  This implies that, theoretically speaking, the performance of the 2D network is the lower bound of the performance of the 3D-2D network for image segmentation tasks.  ... 
doi:10.1186/s12880-020-00509-9 pmid:32972374 fatcat:xuxyr2nqebb37ofj6gmeocnmqe

Attention‐based video object segmentation algorithm

Ying Cao, Lijuan Sun, Chong Han, Jian Guo
2021 IET Image Processing  
To improve the segmentation performance on videos with large object motion or deformation, a novel scheme is proposed which has two branches.  ...  At last, based on the fused results of two branches, the global thresholds and noising area clean method are employed to obtain segmented objects.  ...  Compared with 3D CNN-LSTM-2, the proposed model presents superior performance. In addition, Figure 6 presents a part of FoM results of each videos in CDnet2014.  ... 
doi:10.1049/ipr2.12135 fatcat:gz3poxb3oba5dlnrjblvgey73i

Supervoxel-based segmentation of 3D imagery with optical flow integration for spatiotemporal processing

Xiaohui Huang, Chengliang Yang, Sanjay Ranka, Anand Rangarajan
2018 IPSJ Transactions on Computer Vision and Applications  
UCM performs well on a range of 2D tasks. Consequently, it is somewhat surprising that no 3D version of UCM exists at the present time.  ...  For the latter comparisons, we present results of 3D UCM with and without optical flow video pre-processing.  ...  We also did experiments similar to the pipeline of [67] using 3D UCM for video object segmentation on the Densely Annotated VIdeo Segmentation (DAVIS) dataset.  ... 
doi:10.1186/s41074-018-0045-8 fatcat:syvhbq6vebdf7iveop6hhssx6i

Motion Segmentation and Retrieval for 3D Video Based on Modified Shape Distribution

Toshihiko Yamasaki, Kiyoharu Aizawa
2006 EURASIP Journal on Advances in Signal Processing  
The experimental results using 3D video sequences of dances have demonstrated very promising results for motion segmentation and retrieval.  ...  A similar motion search and retrieval system for 3D video are presented based on a modified shape distribution algorithm. 3D video is a sequence of 3D models made for a real-world object.  ...  In contrast to motion segmentation and retrieval for 3D motion capture data, those for 3D video are much more challenging.  ... 
doi:10.1155/2007/59535 fatcat:okrpxpw2pjfz7pjxmtd46zkz7q
« Previous Showing results 1 — 15 out of 109,407 results