Filters








3,028 Hits in 11.0 sec

Simultaneous Motion Detection and Background Reconstruction with a Conditional Mixed-State Markov Random Field

Tomás Crivelli, Patrick Bouthemy, Bruno Cernuschi-Frías, Jian-feng Yao
2011 International Journal of Computer Vision  
Then (x, y) is said to be a mixed-state conditional random field if x conditioned on y is a mixed-state Markov random field.  ...  How can we formulate a mixed-state Markov random field in order to include continuous and symbolic states within a single random field model?  ... 
doi:10.1007/s11263-011-0429-z fatcat:fm7rk6b4szeslnubgdzb2p2ojm

Inferring ancestral states without assuming neutrality or gradualism using a stable model of continuous character evolution

Michael G Elliot, Arne Ø Mooers
2014 BMC Evolutionary Biology  
Results: We describe Markov chain Monte Carlo methods for fitting the model to biological data paying special attention to ancestral state reconstruction, and study the performance of the model in comparison  ...  The value of a continuous character evolving on a phylogenetic tree is commonly modelled as the location of a particle moving under one-dimensional Brownian motion with constant rate.  ...  BPIC, and members of the UBC and SFU evolution groups for their advice on an early version of this manuscript.  ... 
doi:10.1186/s12862-014-0226-8 pmid:25427971 pmcid:PMC4266906 fatcat:l5l24irswfadjc4oa32dyabome

Motion Textures: Modeling, Classification, and Segmentation Using Mixed-State Markov Random Fields

Tomás Crivelli, Bruno Cernuschi-Frias, Patrick Bouthemy, Jian-Feng Yao
2013 SIAM Journal of Imaging Sciences  
We thus develop a mixed-state Markov random field model to represent motion textures.  ...  First, we present a method for recognition and classification of motion textures, by means of the Kullback-Leibler distance between mixed-state statistical models.  ...  This enables us to solve simultaneous decision-estimation problems, such as motion detection and background reconstruction [21] , in a unified way. Motion textures.  ... 
doi:10.1137/120872048 fatcat:hyuv4jwqxzdgzjsyq4okrpae4y

Relaxing Symmetric Multiple Windows Stereo Using Markov Random Fields [chapter]

Andrea Fusiello, Umberto Castellani, Vittorio Murino
2001 Lecture Notes in Computer Science  
The main aspect is the introduction of a Markov Random Field (MRF) model in the Symmetric Multiple Windows (SMW) stereo algorithm in order to obtain a non-deterministic relaxation.  ...  Results with both synthetic and real stereo pairs demonstrate the improvement over the original SMW algorithm, which was already proven to perform better than state-of-the-art algorithms.  ...  Acknowledgements The authors thank Daniele Zini who wrote the implementation in C of the Zabih and Woodfill algorithm.  ... 
doi:10.1007/3-540-44745-8_7 fatcat:ywchvc22izhknnf75rds2zdvli

Guest Editors' Introduction to the Special Issue on Multimodal Human Pose Recovery and Behavior Analysis

Sergio Escalera, Jordi Gonzalez, Xavier Baro, Jamie Shotton
2016 IEEE Transactions on Pattern Analysis and Machine Intelligence  
In essence, HuPBA requires dealing with the articulated nature of the human body, changes in appearance due to clothing, and the inherent problems of clutter scenes, such as background artifacts, occlusions  ...  recognition, and driver assistance technology, to mention just a few.  ...  Bar o is with the Universitat Oberta de Catalunya and the Computer Vision Center, Catalonia, Spain. E-mail: xbaro@uoc.edu. J. Shotton is with Microsoft Research, Cambridge, United Kingdom.  ... 
doi:10.1109/tpami.2016.2557878 fatcat:ee3j7nre4fgdtjrozavgexvhi4

A Layered Approach for Robust Spatial Virtual Human Pose Reconstruction Using a Still Image

Chengyu Guo, Songsong Ruan, Xiaohui Liang, Qinping Zhao
2016 Sensors  
In this framework, a hierarchical parts-based deep model is used to detect body parts by using the available restricted information in a still image and is then combined with spatial Markov random fields  ...  Pedestrian detection and human pose estimation are instructive for reconstructing a three-dimensional scenario and for robot navigation, particularly when large amounts of vision data are captured using  ...  Author Contributions: Chengyu Guo, Xiaohui Liang and Qingping Zhao conceived of and designed the study.  ... 
doi:10.3390/s16020263 pmid:26907289 pmcid:PMC4801639 fatcat:rvjnuzkxonb5pmsqvwglb22jdu

An Entropic Estimator for Structure Discovery

Matthew Brand
1998 Neural Information Processing Systems  
We introduce a novel framework for simultaneous structure and parameter learning in hidden-variable conditional probability models, based on an en tropic prior and a solution for its maximum a posteriori  ...  Applied to hidden Markov models (HMMs), it finds a concise finite-state machine representing the hidden structure of a signal.  ...  Prior to training, HMM states were initialized to tile the image with their receptive fields, and transition probabilities were initialized to prefer motion to adjoining tiles.  ... 
dblp:conf/nips/Brand98 fatcat:nvlwgbjoszhvfhyt4k3u5huaay

Traditional and recent approaches in background modeling for foreground detection: An overview

Thierry Bouwmans
2014 Computer Science Review  
A B S T R A C T Background modeling for foreground detection is often used in different applications to model the background and then detect the moving objects in the scene like in video surveillance.  ...  Furthermore, we present the available resources, datasets and libraries. Then, we conclude with several promising directions for future research.  ...  Furthermore, the background and foreground models are used competitively in a MAP-MRF (Maximum A Posteriori Markov Random Field) [192] decision framework.  ... 
doi:10.1016/j.cosrev.2014.04.001 fatcat:wccwuwltk5fr7lsgmsu5qbxclm

Author Index

2010 2010 IEEE Computer Society Conference on Computer Vision and Pattern Recognition  
Liao, Shu A Novel Markov Random Field Based Deformable Model for Face Recognition Liebelt, Joerg Multi-View Object Class Detection with a 3D Geometric Model Lim, Jongwoo Egomotion using Assorted  ...  Discrete Minimum Ratio Curves and Surfaces Efficient Piecewise Learning for Conditional Random Fields Torralba, Antonio Exploiting Hierarchical Context on a Large Database of Object Categories Part and  ... 
doi:10.1109/cvpr.2010.5539913 fatcat:y6m5knstrzfyfin6jzusc42p54

Fixed point probability field for complex occlusion handling

F. Fleuret, R. Lengagne, P. Fua
2005 Tenth IEEE International Conference on Computer Vision (ICCV'05) Volume 1  
We start from occupancy probability estimates in a top view and rely on a generative model to yield probability images to be compared with the actual input images.  ...  simple blob detector, and the number of present individuals is a priori unknown.  ...  Most of them strongly rely on temporal information and use a Bayesian framework such as Hidden Markov Models (HMMs) to combine a motion model, that is a probability distribution of state transitions over  ... 
doi:10.1109/iccv.2005.102 dblp:conf/iccv/FleuretLF05 fatcat:2wnr4zev7vhujbqoujqbyavs4a

Energy-based Models for Video Anomaly Detection [article]

Hung Vu, Dinh Phung, Tu Dinh Nguyen, Anthony Trevors, Svetha Venkatesh
2017 arXiv   pre-print
We demonstrate our proposal on the specific application of video anomaly detection and the experimental results indicate that our method performs better than baselines and are comparable with state-of-the-art  ...  Unlike existing appoaches which only partially solve these problems, we develop a unique framework to cope the problems above simultaneously.  ...  A Markov random field with hidden variables is also applied in (Kim and Grauman, 2009) where mixture of probabilistic principle component analysers (MPPCA) is learned on optical flow-based features of  ... 
arXiv:1708.05211v1 fatcat:fkb2m2vdx5fs5grz3mbbmlyctu

Bayesian algorithms for simultaneous structure from motion estimation of multiple independently moving objects

Gang Qian, R. Chellappa, Qinfen Zheng
2005 IEEE Transactions on Image Processing  
By using the proposed algorithms, the relative motions of all moving objects with respect to the camera can be simultaneously estimated.  ...  In this paper, the problem of simultaneous structure from motion estimation for multiple independently moving objects from a monocular image sequence is addressed.  ...  It can be The observed feature trajectories and the mixed motion estimation results using a synthetic sequence with three moving objects. (a) shows feature trajectories in the synthetic sequence.  ... 
doi:10.1109/tip.2004.837551 pmid:15646875 fatcat:s2vd2edf5bdl3ezr2oalxnu7ty

Bayesian estimation of motion vector fields

J. Konrad, E. Dubois
1992 IEEE Transactions on Pattern Analysis and Machine Intelligence  
More recently, Bouthemy and Lalande [5] and Heitz and Bouthemy [ 141, [15] applied Markov random field models to motion detection and segmentation and solved the problem using Besag's ICM method [  ...  The approach stems from well-known concepts used in stochastic modeling and reconstruction of images such as Bayesian estimation criteria [lo] , [27] , Markov random field (MRF) models for images [35  ... 
doi:10.1109/34.161350 fatcat:hk3y5wqqkvafzpfn5phblnl33a

Incorporating visual knowledge representation in stereo reconstruction

A. Barbu, Song-Chun Zhu
2005 Tenth IEEE International Conference on Computer Vision (ICCV'05) Volume 1  
, and a structureless part represented by Markov random field on pixels.  ...  We propose an MCMC algorithm that simultaneously infers the 3D primitive types and parameters and estimates the depth of the scene.  ...  This Markov Random Field together with the labeling of the edges can be thought of as a Mixed Markov Model [6] , in which the neighborhood structure of the MRF depends upon the primitive types, and changes  ... 
doi:10.1109/iccv.2005.120 dblp:conf/iccv/BarbuZ05 fatcat:45w3ldz3frc7rh3uhtpd33lfoe

Artificial Intelligence in Surgery [article]

Xiao-Yun Zhou, Yao Guo, Mali Shen, Guang-Zhong Yang
2019 arXiv   pre-print
We end with summarizing the current state, emerging trends and major challenges in the future development of AI in surgery.  ...  Artificial Intelligence (AI) is gradually changing the practice of surgery with the advanced technological development of imaging, navigation and robotic intervention.  ...  For surgical subtask recognition, most previous works [93, 97, 98] were developed towards variations on Hidden Markov Model (HMM), Conditional Random Field (CRF), and Linear Dy- namic Systems (LDS  ... 
arXiv:2001.00627v1 fatcat:dywtv6v36rgf3fummidyluy3zi
« Previous Showing results 1 — 15 out of 3,028 results