Filters








39,956 Hits in 3.2 sec

Parameterized Modeling and Recognition of Activities

Yaser Yacoob, Michael J. Black
1999 Computer Vision and Image Understanding  
R e c o g n i t i o n o f spatio-temporal varia n t s of modeled activities is achieved b y p a r a m et e r i z i n g t h e ,search in t h e space of admissible transf o r m a t i o n s t h a t t h e  ...  A f r a m e w o r k f o r m o d e l i n g a n d recognition o f t e mporal activities is proposed.  ...  but did not demonstrate modeling and recognition of activities.  ... 
doi:10.1006/cviu.1998.0726 fatcat:xmmtucbovnd3pk4c26im3xinfi

Parameterized modeling and recognition of activities

Y. Yacoob, M.J. Black
Sixth International Conference on Computer Vision (IEEE Cat. No.98CH36271)  
R e c o g n i t i o n o f spatio-temporal varia n t s of modeled activities is achieved b y p a r a m et e r i z i n g t h e ,search in t h e space of admissible transf o r m a t i o n s t h a t t h e  ...  A f r a m e w o r k f o r m o d e l i n g a n d recognition o f t e mporal activities is proposed.  ...  but did not demonstrate modeling and recognition of activities.  ... 
doi:10.1109/iccv.1998.710709 dblp:conf/iccv/YacoobB98 fatcat:hhxrh2ovczc45fir6ora7y5twu

Recognizing Human Motion Using Parameterized Models of Optical Flow [chapter]

Michael J. Black, Yaser Yacoob, Shanon X. Ju
1997 Computational Imaging and Vision  
In this chapter we describe the representation and recognition of human motion using parameterized models of optical ow.  ...  An advantage of the 2D parameterized ow models is that recovered ow parameters can be interprated and used for recognition as described in 7 .  ... 
doi:10.1007/978-94-015-8935-2_11 fatcat:v52kzqzzg5fufnkq6afucqtwke

A probabilistic model with parsinomious representation for sensor fusion in recognizing activity in pervasive environment

D.T. Tran, D.Q. Phung
2006 18th International Conference on Pattern Recognition (ICPR'06)  
., Bui, Hung H. and Venkatesh, Svetha 2006, A probabilistic model with parsinomious representation for sensor fusion in recognizing activity in pervasive environment, in ICPR 2006 : Abstract To tackle  ...  The model is evaluated on real-world data acquired through ubiquitous sensors in recognizing daily morning activities.  ...  In the problem of activity recognition, the higher the accuracy rate and the lower the early detection, the better the recognition performance of the model.  ... 
doi:10.1109/icpr.2006.154 dblp:conf/icpr/TranP06 fatcat:k4o5ckdtn5gdhds3pfmizx4cim

On Human Action [chapter]

Aaron Bobick, Volker Krüger
2011 Visual Analysis of Humans  
HMMs model activity by presuming activity is a first order Markov process and sequence is output from the b j (x). States are hidden and unknown.  Train via expectation/maximization.  ...   a ij is P(q t = j | q t-1 = i)  b j (x) is p(x t = x | q t = j)  Anatomy of Hidden Markov Models C B A Wins and Losses of HMMs in Gesture for Recognition (we never thought about generation  ... 
doi:10.1007/978-0-85729-997-0_14 fatcat:7ndc6on2djg4hjdsg7dhx3reje

A Model Based Approach for Expressions Invariant Face Recognition [chapter]

Zahid Riaz, Christoph Mayer, Matthias Wimmer, Michael Beetz, Bernd Radig
2009 Lecture Notes in Computer Science  
The approach follows in 1) modeling an active appearance model (AAM) parameters for the face image, 2) using optical flow based temporal features for facial expression variations estimation, 3) and finally  ...  The novelty lies not only in generation of appearance models which is obtained by fitting active shape model (ASM) to the face image using objective functions but also using a feature vector which is the  ...  This type of model using shape and texture parameters is called Active Appearance Models (AAMs), introduced by Edwards and Cootes [6] .  ... 
doi:10.1007/978-3-642-01793-3_30 fatcat:5mb7lfoqjnfprmoydgzvcintsq

Meta-learning of Pooling Layers for Character Recognition [article]

Takato Otsuzuki, Heon Song, Seiichi Uchida, Hideaki Hayashi
2021 arXiv   pre-print
The results demonstrate that a pooling layer that is suitable across character recognition tasks was obtained via meta-learning, and the obtained pooling layer improved the performance of the model in  ...  As part of our framework, a parameterized pooling layer is proposed in which the kernel shape and pooling operation are trainable using two parameters, thereby allowing flexible pooling of the input data  ...  Acknowledgments: This work was supported by JSPS KAKENHI Grant Numbers JP17K12752 and JP17H06100.  ... 
arXiv:2103.09528v2 fatcat:bwlr33mdbfgy7fk7nlpthor3ie

International Conference on Pattern Recognition

1971 Computer  
Recognition tests on a subset of the FRGC data set show approximately 80% rank-one recognition rate using only the eyes and nose part of the face.  ...  In this paper, we introduce a new approach for partial 3D face recognition, which makes use of shape decomposition over the rigid 1 part of a face.  ...  From a practical point of view, the conformal parameterization is easy to control. Hence conformal parameterization is crucial for 3D shape matching and recognition.  ... 
doi:10.1109/c-m.1971.216793 fatcat:ghskjx6ylbhkflhg4meuskb3cy

Efficient Coxian Duration Modelling for Activity Recognition in Smart Environments with the Hidden semi-Markov Model

T.V. Duong, D.Q. Phung, H.H. Bui, S. Venkatesh
2005 2005 International Conference on Intelligent Sensors, Sensor Networks and Information Processing  
S. 2005, Efficient Coxian duration modelling for activity recognition in smart environment with the hidden semi-Markov model, in Proceedings of the 2005 intelligent sensors, sensor networks and information  ...  Abstract In this paper, we exploit the discrete Coxian distribution and propose a novel form of stochastic model, termed as the Coxian hidden semi-Makov model (Cox-HSMM), and apply it to the task of recognising  ...  This form of stochastic model has been an active research topic since the late 1980s driven mainly in the field of speech processing and recognition [3] , [11] .  ... 
doi:10.1109/issnip.2005.1595592 fatcat:lk5arah5b5dpxe4andhtfq32ce

Partial Face Biometry Using Shape Decomposition on 2D Conformal Maps of Faces

Przemyslaw Szeptycki, Mohsen Ardabilian, Liming Chen, Wei Zeng, David Gu, Dimitris Samaras
2010 2010 20th International Conference on Pattern Recognition  
Recognition tests on a subset of the FRGC data set show approximately 80% rank-one recognition rate using only the eyes and nose part of the face.  ...  In this paper, we introduce a new approach for partial 3D face recognition, which makes use of shape decomposition over the rigid 1 part of a face.  ...  From a practical point of view, the conformal parameterization is easy to control. Hence conformal parameterization is crucial for 3D shape matching and recognition.  ... 
doi:10.1109/icpr.2010.372 dblp:conf/icpr/SzeptyckiACZGS10 fatcat:yywu2mhdznby7fgdurjf544o2a

MKPLS: Manifold Kernel Partial Least Squares for Lipreading and Speaker Identification

Amr Bakry, Ahmed Elgammal
2013 2013 IEEE Conference on Computer Vision and Pattern Recognition  
We show the results for three different settings of lipreading: speaker independent, speaker dependent, and speaker semi-dependent.  ...  Our approach outperforms for the speaker semi-dependent setting by at least 15% of the baseline, and competes in the other two settings.  ...  This work was also partly supported by the Office of Navel Research grant N00014-12-1-0755.  ... 
doi:10.1109/cvpr.2013.94 dblp:conf/cvpr/BakryE13 fatcat:ecbp5gt52jbdxcaus5qxjahreu

Manifold-Kernels Comparison in MKPLS for Visual Speech Recognition [article]

Amr Bakry, Ahmed Elgammal
2016 arXiv   pre-print
This work is intended to evaluate the performance of several manifold kernels for solving the problem of visual speech recognition. We show the theory behind each kernel.  ...  One common approach is to model the visual recognition as nonlinear optimization problem. Measuring the distances between visual units is essential for solving this problem.  ...  In [7] , HMM was used for encoding the visual dynamics of speech using Active Shape Model (ASM) and Active Appearance Model (AAM).  ... 
arXiv:1601.05861v1 fatcat:5sftjcgxhrezza7vu6se3szlqa

Personalizing Gesture Recognition Using Hierarchical Bayesian Neural Networks

Ajjen Joshi, Soumya Ghosh, Margrit Betke, Stan Sclaroff, Hanspeter Pfister
2017 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)  
We also develop methods for adapting our model to new subjects when a small number of subject-specific personalization data is available.  ...  Focusing on the problem of gesture recognition where inter-subject variations are commonplace, we demonstrate the effectiveness of our proposed techniques.  ...  The authors wish to thank Jessica Hodgins, Leonid Sigal, Scott Watson, Jamie Robertson, and Michael Holton. This work was supported in part by Disney Research and NSF grants 1551572 and 1337866.  ... 
doi:10.1109/cvpr.2017.56 dblp:conf/cvpr/JoshiGBSP17 fatcat:5k6dnxplrvftfh7ctelxv3sxte

Parametric hidden Markov models for gesture recognition

A.D. Wilson, A.F. Bobick
1999 IEEE Transactions on Pattern Analysis and Machine Intelligence  
AbstractÐA new method for the representation, recognition, and interpretation of parameterized gesture is presented.  ...  The nonlinear formulation requires the use of a generalized expectation-maximization (GEM) algorithm for both training and the simultaneous recognition of the gesture and estimation of the value of the  ...  MOTIVATION AND PRIOR WORK Using HMMs in Gesture Recognition Hidden Markov models and related techniques have been applied to gesture recognition tasks with success.  ... 
doi:10.1109/34.790429 fatcat:eopsgyqinfh7rn2waptrxgt5ly

Digit Recognition Based on Specialization, Decomposition and Holistic Processing

Michael Joseph, Khaled Elleithy
2020 Machine Learning and Knowledge Extraction  
With the introduction of the Convolutional Neural Network (CNN) and other classical algorithms, facial and object recognition have made significant progress.  ...  The model uses a variational autoencoder to generate holistic representation of handwritten digits and a Neural Network(NN) to classify them.  ...  The cost of publishing this paper was funded by the University of Bridgeport, CT, USA. Conflicts of Interest: The authors declare no conflict of interest.  ... 
doi:10.3390/make2030015 fatcat:ht4s2nftizbdnekfyha7iest54
« Previous Showing results 1 — 15 out of 39,956 results