Filters








11,380 Hits in 7.7 sec

Graph-Based Correlated Topic Model for Trajectory Clustering in Crowded Videos

Manal Al Ghamdi, Yoshihiko Gotoh
2018 2018 IEEE Winter Conference on Applications of Computer Vision (WACV)  
Unlike previous works that depend on scenes prior, we extract trajectories and apply a spatio-temporal graph (STG) to uncover the spatial and temporal coherence between the trajectories during the learning  ...  This paper presents a graph-based correlated topic model (GCTM) to learn and analyse motion patterns by trajectory clustering in a highly cluttered and crowded environment.  ...  Many works have been proposed for trajectory clustering based on mid-level features learning.  ... 
doi:10.1109/wacv.2018.00118 dblp:conf/wacv/GhamdiG18 fatcat:v6khgdcde5gijbcew3hi75o2nm

Progressive Dictionary Learning With Hierarchical Predictive Structure for Low Bit-Rate Scalable Video Coding

Wenrui Dai, Yangmei Shen, Hongkai Xiong, Xiaoqian Jiang, Junni Zou, David Taubman
2017 IEEE Transactions on Image Processing  
On the other hand, scalable extension of conventional hybrid motion compensation-discrete cosine transform (MC-DCT) framework has been widely studied for scalable video coding.  ...  This paper proposes a progressive dictionary learning framework with hierarchical predictive structure for scalable video coding, especially in low bitrate region.  ...  The proposed framework for scalable video coding based on progressive dictionary learning.  ... 
doi:10.1109/tip.2017.2692882 pmid:28422683 pmcid:PMC5638692 fatcat:yrf6mgcpojezvjhmiuvmvqklne

A Gaussian Mixture Representation of Gesture Kinematics for On-Line Sign Language Video Annotation [chapter]

Fabio Martínez, Antoine Manzanera, Michèle Gouiffès, Annelies Braffort
2015 Lecture Notes in Computer Science  
This work introduces a method to support SL annotations based on a motion descriptor that characterizes dynamic gestures in videos.  ...  Then for each sub-region, a histogram of motion-cues occurrence is built, forming a frame-gesture descriptor which can be used for on-line annotation.  ...  Conclusions This work introduced a new motion descriptor that is able to recognize motion gestures related with SLs. The proposed approach allows an on-line support to SL annotation, by  ... 
doi:10.1007/978-3-319-27863-6_27 fatcat:qyr23t7y2fh7nlhvqmnzgl47nu

Human Action Recognition Using Improved Salient Dense Trajectories

Qingwu Li, Haisu Cheng, Yan Zhou, Guanying Huo
2016 Computational Intelligence and Neuroscience  
In this paper, we present a more effective approach of video representation using improved salient dense trajectories: first, detecting the motion salient region and extracting the dense trajectories by  ...  Human action recognition in videos is a topic of active research in computer vision. Dense trajectory (DT) features were shown to be efficient for representing videos in state-of-the-art approaches.  ...  We first extract the motion salient region in videos.  ... 
doi:10.1155/2016/6750459 pmid:27293425 pmcid:PMC4886085 fatcat:g67kevkmkvdzxngamtivdregae

Action Recognition in Realistic Sports Videos [chapter]

Khurram Soomro, Amir R. Zamir
2014 Advances in Computer Vision and Pattern Recognition  
We study several recent methods for action localization which have shown promising results on sports videos.  ...  To provide further details about the existing action recognition methods in this area, we decompose the action recognition framework into three main steps of feature extraction, dictionary learning to  ...  The majority of existing frameworks for action recognition consist of three main steps: feature extraction, dictionary learning to form a representation for a video based on the extracted features, and  ... 
doi:10.1007/978-3-319-09396-3_9 fatcat:q5kjcsnp2zb7fo5j4wwkjel5re

Unsupervised human activity analysis for intelligent mobile robots

Paul Duckworth, David C. Hogg, Anthony G. Cohn
2019 Artificial Intelligence  
In this thesis an approach for unsupervised learning of activities implemented on an autonomous mobile robot is presented.  ...  Finally, we present methods for learning such human activity models in an incremental and continuous setting using variational inference methods to update the activity distribution online.  ...  Pixel-based abstraction This abstraction method relies on pixel-based information extracted from images such as colour, texture or gradients.  ... 
doi:10.1016/j.artint.2018.12.005 fatcat:vq56rrsuojfq3kzr3ip3tj4d74

Trajectory clustering for motion pattern extraction in aerial videos

Tahir Nawaz, Andrea Cavallaro, Bernhard Rinner
2014 2014 IEEE International Conference on Image Processing (ICIP)  
We present an end-to-end approach for trajectory clustering from aerial videos that enables the extraction of motion patterns in urban scenes.  ...  Then clustering is performed based on statistics from the Discrete Wavelet Transform coefficients extracted from the trajectories.  ...  This paper presents an end-to-end approach for trajectory clustering for motion pattern extraction in aerial videos.  ... 
doi:10.1109/icip.2014.7025203 dblp:conf/icip/NawazCR14 fatcat:jmjcoxr3xrhm5eto6uf2tw6g5q

Unsupervised Learning of Long-Term Motion Dynamics for Videos

Zelun Luo, Boya Peng, De-An Huang, Alexandre Alahi, Li Fei-Fei
2017 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)  
We argue that in order for the decoder to reconstruct these sequences, the encoder must learn a robust video representation that captures long-term motion dependencies and spatial-temporal relations.  ...  Given a pair of images from a video clip, our framework learns to predict the long-term 3D motions.  ...  Next, we specially thank Juan Carlos Niebles, Serena Yeung, Kenji Hata, Yuliang Zou, and Lyne Tchapmi for their helpful feedback.  ... 
doi:10.1109/cvpr.2017.751 dblp:conf/cvpr/LuoPHAF17 fatcat:mnk4ejnkcreg5edgogrmoq6nuy

Learning Temporal Regularity in Video Sequences

Mahmudul Hasan, Jonghyun Choi, Jan Neumann, Amit K. Roy-Chowdhury, Larry S. Davis
2016 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)  
We approach this problem by learning a generative model for regular motion patterns (termed as regularity) using multiple sources with very limited supervision.  ...  We evaluate our methods in both qualitative and quantitative ways -showing the learned regularity of videos in various aspects and demonstrating competitive performance on anomaly detection datasets as  ...  Most video based anomaly detection approaches involve a local feature extraction step followed by learning a model on training video.  ... 
doi:10.1109/cvpr.2016.86 dblp:conf/cvpr/0003CNRD16 fatcat:z4gq5mndhrcrbmitar2ldywfvq

Towards Good Practices for Action Video Encoding

Jianxin Wu, Yu Zhang, Weiyao Lin
2014 2014 IEEE Conference on Computer Vision and Pattern Recognition  
We empirically evaluated various VLAD improvement technologies to determine good practices in VLAD-based video encoding.  ...  Furthermore, we propose an interpretation that VLAD is a maximum entropy linear feature learning process.  ...  Thus, Table 3 indicates that on average, more than 15% code words are missing in any single video-a fact that carries useful side information for categorizing a video's action type.  ... 
doi:10.1109/cvpr.2014.330 dblp:conf/cvpr/WuZL14 fatcat:6xy4uu5nizbubdtdc3dcodrmau

Learning to Manipulate Tools by Aligning Simulation to Video Demonstration [article]

Kateryna Zorina, Justin Carpentier, Josef Sivic, Vladimír Petrík
2021 arXiv   pre-print
For this we combine reinforcement learning with an optimization procedure to find a control policy and the placement of the robot based on the tool motion in the aligned environment.  ...  Second, we describe an imitation learning approach that focuses on the trajectory of the tool rather than the motion of the human.  ...  sand and the target box) based on the trajectory extracted from the video; (ii) we sample the simulation parameters from the constructed distribution, (iii) we follow the extracted trajectory in simulated  ... 
arXiv:2111.03088v1 fatcat:2i6sdoddyrgqtagoyfqqdmat44

Sports Video Analysis: Semantics Extraction, Editorial Content Creation and Adaptation

Changsheng Xu, Jian Cheng, Yi Zhang, Yifan Zhang, Hanqing Lu
2009 Journal of Multimedia  
In this paper, we summarize our research achievement on semantics extraction and automatic editorial content creation and adaptation in sports video analysis.  ...  We first propose a generic multi-layer and multi-modal framework for sports video analysis.  ...  On the server side, the highlights of the sports video were extracted based on replay detection.  ... 
doi:10.4304/jmm.4.2.69-79 fatcat:xytusontr5cyxlxpyqgljnhkqu

Video Activity Recognition: State-of-the-Art

Itsaso Rodríguez-Moreno, José María Martínez-Otzeta, Basilio Sierra, Igor Rodriguez, Ekaitz Jauregi
2019 Sensors  
In the area of robotics, the tasks of autonomous navigation or social interaction could also take advantage of the knowledge extracted from live video recording.  ...  The aim of this paper is to survey the state-of-the-art techniques for video activity recognition while at the same time mentioning other techniques used for the same task that the research community has  ...  One of the main techniques used for activity recognition is computer vision, namely video-based activity recognition. Visual video features provide basic information for video events or actions.  ... 
doi:10.3390/s19143160 fatcat:ore5paemmbfxzjimsvnnezgwwi

Efficient Analysis of Traffic Intersection Scenes by Employing Traffic Phase Information

P. Ahmadi, I. Gholampour
2019 Iranian Journal of Electrical and Electronic Engineering  
Using side information on traffic phases, the semantic of motion patterns from traffic intersection scenes can be learned more effectively.  ...  The learning is performed based on optical flow features extracted from training video clips, and applying them to supervised topic models such as MedLDA and MedSTC.  ...  Based on such auxiliary information, we have formulated the problem of traffic motion pattern extraction and abnormality detection under supervised topic modeling framework.  ... 
doaj:5d2e4449c10a45fd895a32ab55d8570f fatcat:ckqvaduvx5fdjfskhc3rmz4cui

Regularity Learning via Explicit Distribution Modeling for Skeletal Video Anomaly Detection [article]

Shoubin Yu, Zhongyin Zhao, Haoshu Fang, Andong Deng, Haisheng Su, Dongliang Wang, Weihao Gan, Cewu Lu, Wei Wu
2021 arXiv   pre-print
These two modules are then integrated into a unified framework for pose regularity learning, which is referred to as Motion Prior Regularity Learner (MoPRL).  ...  Anomaly detection in surveillance videos is challenging and important for ensuring public security.  ...  Specifically, we use the tools [12, 36] to obtain the pose trajectories as in [25] on ShanghaiTech. While for the Corridor dataset, we extract pose trajectories with tools [3, 7] as in [30] .  ... 
arXiv:2112.03649v2 fatcat:wgdridft3zaenls7kts5badtbq
« Previous Showing results 1 — 15 out of 11,380 results