Filters








1,982 Hits in 8.6 sec

Temporally Consistent Motion Segmentation from RGB-D Video [article]

Peter Bertholet, Alexandru-Eugen Ichim, Matthias Zwicker
2016 arXiv   pre-print
We present a method for temporally consistent motion segmentation from RGB-D videos assuming a piecewise rigid motion model.  ...  We develop a novel initialization procedure that clusters feature tracks obtained from the RGB data by leveraging the depth information.  ...  Conclusions We presented a novel method for temporally consistent motion segmentation from RGB-D videos.  ... 
arXiv:1608.04642v1 fatcat:2ec3jblhwnh4bo5yekc6vkgi44

Human Fall Detection Based on Body Posture Spatio-Temporal Evolution

Jin Zhang, Cheng Wu, Yiming Wang
2020 Sensors  
Based on the study of the stability of human body dynamics, the article proposes a new model of human posture representation of fall behavior, called the "five-point inverted pendulum model", and uses  ...  Vision-driven fall event detection has the huge advantage of being non-invasive. However, in actual scenes, the fall behavior is rich in diversity, resulting in strong instability in detection.  ...  [27] proposes a novel efficient high-level representation of trajectories called snapped trajectories, and combines object trajectory analysis and pixel based analysis for abnormal behavior inference  ... 
doi:10.3390/s20030946 pmid:32050727 pmcid:PMC7039221 fatcat:w4vc6gycbvh3jcs6w77wf2fysi

Integrating depth and color cues for dense multi-resolution scene mapping using RGB-D cameras

Jorg Stuckler, Sven Behnke
2012 2012 IEEE International Conference on Multisensor Fusion and Integration for Intelligent Systems (MFI)  
The mapping of environments is a prerequisite for many navigation and manipulation tasks. We propose a novel method for acquiring 3D maps of indoor scenes from a freely moving RGB-D camera.  ...  For simultaneous localization and mapping, we extract key views and optimize the trajectory in a probabilistic framework.  ...  We propose a novel approach to acquire explicit geometric models of indoor scenes with a RGB-D camera that include color and depth cues seamlessly.  ... 
doi:10.1109/mfi.2012.6343050 dblp:conf/mfi/StucklerB12 fatcat:q7ienagdvre6zbrrfwv22z7cxa

Human Action Recognition and Prediction: A Survey [article]

Yu Kong, Yun Fu
2022 arXiv   pre-print
Many attempts have been devoted in the last a few decades in order to build a robust and effective framework for action recognition and prediction.  ...  Derived from rapid advances in computer vision and machine learning, video analysis tasks have been moving from inferring the present state to predicting the future state.  ...  Classifiers for RGB-D Videos Action recognition from RGB-D videos has been receiving a lot of attentions [270, 271, 75, 294, 156, 190] due to the advent of the cost-effective Kinect sensor [232] .  ... 
arXiv:1806.11230v3 fatcat:2a2d7fuezbdqzfgrjwkcuqvmbu

A survey on vision-based fall detection

Zhong Zhang, Christopher Conly, Vassilis Athitsos
2015 Proceedings of the 8th ACM International Conference on PErvasive Technologies Related to Assistive Environments - PETRA '15  
Falls are a major cause of fatal injury for the elderly population.  ...  To improve the quality of living for seniors, a wide range of monitoring systems with fall detection functionality have been proposed over recent years.  ...  [39] propose a RGB-D cameras based method to recognize five activities: standing, falling from standing, falling from chair, sitting on chair, and sitting on floor.  ... 
doi:10.1145/2769493.2769540 dblp:conf/petra/ZhangCA15 fatcat:f3gjqvybojalla7igfa2tvz4yy

A Review on Human Activity Recognition Using Vision-Based Method

Shugang Zhang, Zhiqiang Wei, Jie Nie, Lei Huang, Shuang Wang, Zhen Li
2017 Journal of Healthcare Engineering  
For the representation methods, we sort out a chronological research trajectory from global representations to local representations, and recent depth-based representations.  ...  For the classification methods, we conform to the categorization of template-based methods, discriminative models, and generative models and review several prevalent methods.  ...  Shandong Province (no. 2015ZDZX05002); Qingdao Science and Technology Development Plan (no. 16-5-1-13-jch); and The Aoshan Innovation Project in Science and Technology of Qingdao National Laboratory for  ... 
doi:10.1155/2017/3090343 pmid:29065585 pmcid:PMC5541824 fatcat:g6qbbbjpcref3p54kvquu5rltq

3D Tracker-Level Fusion for Robust RGB-D Tracking

Ning AN, Xiao-Guang ZHAO, Zeng-Guang HOU
2017 IEICE transactions on information and systems  
We perform extensive experiments on a large-scale RGB-D benchmark dataset.  ...  The evaluation results demonstrate the effectiveness of the proposed fusion algorithm and the superior performance of the proposed TLF3D tracker against state-of-the-art RGB-D trackers. key words: RGB-D  ...  Conclusion In this study, we propose a novel 3D tracker-level fusion algorithm (TLF3D) for online RGB-D tracking, where fusion is performed by optimizing both the 3D cube attraction energy and the 3D trajectory  ... 
doi:10.1587/transinf.2016edp7498 fatcat:vxwiiiidqnazrlm3lwtf6wfxjq

Learning Articulated Motions From Visual Demonstration

Sudeep Pillai, Matthew Walter, Seth Teller
2014 Robotics: Science and Systems X  
Using the method, a robot can observe an object being exercised, infer a kinematic model incorporating rigid, prismatic and revolute joints, then use the model to predict the object's motion from a novel  ...  A robotic mobile manipulator would benefit from the ability to acquire kinematic models of such objects from observation.  ...  For 3-D key-points, we use γ = 1 2 cm in Eqn. 1.  ... 
doi:10.15607/rss.2014.x.050 dblp:conf/rss/PillaiWT14 fatcat:d7xnxgiezzg2jivugrx6dic3yi

Learning Articulated Motions From Visual Demonstration [article]

Sudeep Pillai, Matthew R. Walter, Seth Teller
2015 arXiv   pre-print
Using the method, a robot can observe an object being exercised, infer a kinematic model incorporating rigid, prismatic and revolute joints, then use the model to predict the object's motion from a novel  ...  A robotic mobile manipulator would benefit from the ability to acquire kinematic models of such objects from observation.  ...  For 3-D key-points, we use γ = 1 2 cm in Eqn. 1.  ... 
arXiv:1502.01659v1 fatcat:kg7gecswefes3irre7zlakc3pq

Prediction of Human Activities Based on a New Structure of Skeleton Features and Deep Learning Model

Neziha Jaouedi, Francisco J. Perales, José Maria Buades, Noureddine Boujnah, Med Salim Bouhlel
2020 Sensors  
In this paper, we present a novel approach to human activity recognition from videos using the Recurrent Neural Network (RNN) for activity classification and the Convolutional Neural Network (CNN) with  ...  The performance of the proposed approach is evaluated by the RGB-D sensor dataset CAD-60.  ...  ., 2016 [28] used RGB-D sensors for human skeleton detection and kinetic energy to identify the key poses that present intense positions of action in a large space.  ... 
doi:10.3390/s20174944 pmid:32882884 fatcat:wju6gvbiibchhno5r4c733uloy

RGB-D Data-Based Action Recognition: A Review

Muhammad Bilal Shaikh, Douglas Chai
2021 Sensors  
In this paper, we focus solely on data fusion and recognition techniques in the context of vision with an RGB-D perspective.  ...  Naturally, each action-data modality—such as RGB, depth, skeleton, and infrared (IR)—has distinct characteristics; therefore, it is important to exploit the value of each modality for better action recognition  ...  Acknowledgments: The authors would like to thank the anonymous reviewers for their careful reading and valuable remarks, which have greatly helped extend the scope of this paper.  ... 
doi:10.3390/s21124246 fatcat:7dvocdy63rckne5yunhfsnr4p4

Movable-Object-Aware Visual SLAM via Weakly Supervised Semantic Segmentation [article]

Ting Sun, Yuxiang Sun, Ming Liu, Dit-Yan Yeung
2019 arXiv   pre-print
In this paper we propose to adopt a novel weakly-supervised semantic segmentation method. The segmentation mask is obtained from a CNN pre-trained with image-level class labels only.  ...  Experimental results on the TUM RGB-D and the KITTI stereo datasets demonstrate our superiority over the state-of-the-art.  ...  Trajectories of sequences (from the left column to the right) f3/w/xyz, f3/w/rpy, f3/w/static, and f3/w/halfsphere, f3/s/static from dynamic objects sequences of TUM RGB-D dataset [26] .  ... 
arXiv:1906.03629v2 fatcat:ebynlvgfjfdtth7hnao3guq73a

RGB-Depth SLAM Review [article]

Redhwan Jamiruddin, Ali Osman Sari, Jahanzaib Shabbir, Tarique Anwer
2018 arXiv   pre-print
The algorithms and approaches are compared for their effectiveness in tracking and mapping based on Root Mean Square error over online available datasets.  ...  The frontal end utilizes visual imagery of the RGB-D sensor for detection of the key points as well as extracts descriptors.  ...  The novel method to visualization of SLAM algorithms with the use of RGB-D sensors means the method will entail the methodologies that entail extracting visual key points from colored images and applying  ... 
arXiv:1805.07696v1 fatcat:2hee45pltfgi7k262yh6qx5e2e

Object Discovery in Videos as Foreground Motion Clustering

Christopher Xie, Yu Xiang, Zaid Harchaoui, Dieter Fox
2019 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)  
We introduce a novel pixel-trajectory recurrent neural network that learns feature embeddings of foreground pixel trajectories linked across time.  ...  We consider the problem of providing dense segmentation masks for object discovery in videos.  ...  Acknowledgements We thank Pia Bideau for providing evaluation code.  ... 
doi:10.1109/cvpr.2019.01023 dblp:conf/cvpr/XieXHF19 fatcat:oelz4rcggjgypn6wrcehpsjhhe

Object Discovery in Videos as Foreground Motion Clustering [article]

Christopher Xie, Yu Xiang, Zaid Harchaoui, Dieter Fox
2019 arXiv   pre-print
We introduce a novel pixel-trajectory recurrent neural network that learns feature embeddings of foreground pixel trajectories linked across time.  ...  We consider the problem of providing dense segmentation masks for object discovery in videos.  ...  Acknowledgements We thank Pia Bideau for providing evaluation code.  ... 
arXiv:1812.02772v2 fatcat:g46hutdufbh5vlgbptsu6k5qoi
« Previous Showing results 1 — 15 out of 1,982 results