Filters








1,637 Hits in 2.2 sec

Visual Tracking via Nonnegative Multiple Coding

Fanghui Liu, Chen Gong, Tao Zhou, Keren Fu, Xiangjian He, Jie Yang
2017 IEEE transactions on multimedia  
To address this problem, in this paper, we propose a novel appearance model named as "Nonnegative Multiple Coding" (NMC) to accurately represent a target.  ...  nonnegative purpose with theoretical guarantees.  ...  Based on the above observations, this paper develops a ro-bust Nonnegative Multiple Coding (NMC) tracker by exploiting an ensemble of multiple dictionaries and the nonnegative constraint to accurately  ... 
doi:10.1109/tmm.2017.2708424 fatcat:4t6gvypvbnhetfmcvxecgc2osq

Local Topic Discovery via Boosted Ensemble of Nonnegative Matrix Factorization

Sangho Suh, Jaegul Choo, Joonseok Lee, Chandan K. Reddy
2017 Proceedings of the Twenty-Sixth International Joint Conference on Artificial Intelligence  
To tackle this problem, we propose a novel ensemble model of nonnegative matrix factorization for discovering high-quality local topics.  ...  Nonnegative matrix factorization (NMF) has been increasingly popular for topic modeling of large-scale documents.  ...  In response, this paper proposes a local ensemble model of nonnegative matrix factorization (NMF) [Lee and Seung, 1999] .  ... 
doi:10.24963/ijcai.2017/699 dblp:conf/ijcai/SuhCLR17 fatcat:djlmx6h2m5e2dhvumzjveheol4

Sparse coding and dimensionality reduction in cortex [article]

Michael Beyeler, Emily Rounds, Kristofor Carlson, Nikil Dutt, Jeffrey L. Krichmar
2017 bioRxiv   pre-print
sparse coding (NSC) by efficiently encoding high-dimensional stimulus spaces using a sparse and parts-based population code.  ...  Supported by recent computational studies, sparse coding and dimensionality reduction are emerging as a ubiquitous coding strategy across brain regions and modalities, allowing neurons to achieve nonnegative  ...  When even and odd trials on the same track locations were compared, ensemble prediction error was very low, but when the tracks were in two different locations (i.e., α vs. β), the prediction error was  ... 
doi:10.1101/149880 fatcat:dj7uczsby5bspdezr5uuc7qb5a

Model Averaging via Penalized Regression for Tracking Concept Drift

Kyupil Yeon, Moon Sup Song, Yongdai Kim, Hosik Choi, Cheolwoo Park
2010 Journal of Computational And Graphical Statistics  
Using various numerical examples, it is shown that the proposed algorithm can track concept drift better than other existing ensemble methods.  ...  We propose a new model combining algorithm for tracking concept drift in data streams. The final predictive ensemble model has a form of a weighted average and ridge regression combiner.  ...  (appendix.pdf) R-package for the proposed algorithm: R-package contains the code to perform the methods described in the article. (drift 0.1.zip)  ... 
doi:10.1198/jcgs.2010.08104 fatcat:222iby7xcnfjxpmv4i6mmujhze

A Detection Framework of Malicious Code Based on Multi-Classifiers Ensemble

Chao Dai, Jianmin Pang, Feng Yue, Pingfei Cui, Di Sun, Liang Zhu
2016 International Journal of Security and Its Applications  
We utilized multi-classifiers ensemble based on fuzzy integral to improve the accuracy of the detection framework.  ...  In view of the situation mentioned above, in this paper we presented a detection framework based on multi-classifiers ensemble.  ...  In the detection framework of malicious code based on multi-classifiers ensemble, the fuzzy integral is corresponding to a chromosome.  ... 
doi:10.14257/ijsia.2016.10.6.09 fatcat:5nzner2juzgqznf4cbnip2vu7a

Learning and storing the parts of objects: IMF

Ruairi de Frein
2014 2014 IEEE International Workshop on Machine Learning for Signal Processing (MLSP)  
INTRODUCTION Nonnegative Matrix Factorization (NMF) uncovers feature vectors and vectors of activations of these features in a signal ensemble with structural regularity via non-negativity constraints  ...  , email analysis. 2) Image analysis and computer vision: Feature representation and sparse coding; Video tracking. 3) Social Network Analysis: Community structure and trend detection; Recommendation systems  ... 
doi:10.1109/mlsp.2014.6958926 dblp:conf/mlsp/Frein14 fatcat:i22ekoaizfdv5ktgxgdnr27atq

Randomized Ensemble Tracking

Qinxun Bai, Zheng Wu, Stan Sclaroff, Margrit Betke, Camille Monnier
2013 2013 IEEE International Conference on Computer Vision  
We propose a randomized ensemble algorithm to model the time-varying appearance of an object for visual tracking.  ...  In contrast with previous online methods for updating classifier ensembles in tracking-by-detection, the weight vector that combines weak classifiers is treated as a random variable and the posterior distribution  ...  This material is based on work supported by the US National Science Foundation under Grant Nos. 0910908 and 0855065 and the US Army under TARDEC STTR Contract No. W56HZV-12-C-0049.  ... 
doi:10.1109/iccv.2013.255 dblp:conf/iccv/BaiWSBM13 fatcat:sraxo262jbf6zdiah233sq6fga

Semi-Nonnegative Matrix Factorization for Motion Segmentation with Missing Data [chapter]

Quanyi Mo, Bruce A. Draper
2012 Lecture Notes in Computer Science  
In this paper, we present a Semi-Nonnegative Matrix Factorization (SNMF)method that models dense point tracks in terms of their optical flow, and decomposes sets of point tracks into semantically meaningful  ...  Our first contribution is a motion segmentation algorithm based on Semi-Nonnegative Matrix Factorization (SNMF) with missing values.  ...  The non-negative constraint on G is based on the common observation that NMF methods can extract meaningful "parts" of ensemble data [8] [10] .  ... 
doi:10.1007/978-3-642-33786-4_30 fatcat:d7xijnyci5cxjcr7id5p6muv6m

Randomly Punctured LDPC Codes

David G. M. Mitchell, Michael Lentmaier, Ali E. Pusane, Daniel J. Costello
2016 IEEE Journal on Selected Areas in Communications  
Protograph-based LDPC block code and spatially coupled LDPC code ensembles are used throughout as examples to demonstrate the results.  ...  mother code ensemble.  ...  We refer to the range of rates R(0) ≤ R(α) ≤ R max where the punctured code ensembles have nonnegative thresholds as the achievable rate range.  ... 
doi:10.1109/jsac.2015.2507758 fatcat:ke3w4azypnbntppyajmiotdl4i

Windowed Decoding of Protograph-Based LDPC Convolutional Codes Over Erasure Channels

Aravind R. Iyengar, Marco Papaleo, Paul H. Siegel, Jack Keil Wolf, Alessandro Vanelli-Coralli, Giovanni E. Corazza
2012 IEEE Transactions on Information Theory  
We consider a windowed decoding scheme for LDPC convolutional codes that is based on the belief-propagation (BP) algorithm.  ...  We will consider the performance of these ensembles and codes over erasure channels with and without memory.  ...  Since we require for a nonnegative code rate in (2) , which shows that the MBL of an MDS code can never be achieved.  ... 
doi:10.1109/tit.2011.2177439 fatcat:tr4ooffz2ng2bdicahz3334jua

Multiple-view object recognition in band-limited distributed camera networks

Allen Y. Yang, Subhransu Maji, C. Mario Christoudias, Trevor Darrell, Jitendra Malik, S. Shankar Sastry
2009 2009 Third ACM/IEEE International Conference on Distributed Smart Cameras (ICDSC)  
On the base station, we study multiple decoding schemes to simultaneously recover the multiple-view object features based on the distributed compressive sensing theory.  ...  The ability to perform robust object recognition is crucial for applications such as visual surveillance to track and identify objects of interest, and compensate visual nuisances such as occlusion and  ...  Hence, they can be simultaneously recovered on the base computer via a 1 -min solver such as the nonnegative polytope faces pursuit.  ... 
doi:10.1109/icdsc.2009.5289410 dblp:conf/icdsc/YangMCDMS09 fatcat:bix4xkfofjfd5ccd6pjdk4fhka

Local Low Dimensionality and Relation to Effects Of Targeted Weather Observations

D. J. Patil
2003 AIP Conference Proceedings  
To track the propagation of the targeted low dimensional spot, an additional forecast ensemble was created.  ...  The major difference between the two ensembles is that the results presented here are based on an ensemble that uses many (six times) more bred vectors.  ... 
doi:10.1063/1.1612207 fatcat:3vhieodt7ffipjvvu5egsl4ryi

Real-Time Data Driven Wildland Fire Modeling [chapter]

Jonathan D. Beezley, Soham Chakraborty, Janice L. Coen, Craig C. Douglas, Jan Mandel, Anthony Vodacek, Zhen Wang
2008 Lecture Notes in Computer Science  
Data are assimilated using both amplitude and position corrections using a morphing ensemble Kalman filter.  ...  We will use thermal images of a fire for observations that will be compared to synthetic image based on the model state.  ...  Instead of tracking the fire by a custom ad hoc tracer code, it is now represented using a level set method [9] .  ... 
doi:10.1007/978-3-540-69389-5_7 fatcat:isx2wy26vzbzvl3n7eg2fqbs2i

Ultrafast Readout of Representations from Spatially Distributed Rodent Hippocampal Field Potentials [article]

Liang Cao, Viktor Varga, Zhe S Chen
2019 bioRxiv   pre-print
R turns) based on either ensemble spikes or LFP features collected in the central arm of T-maze.  ...  Real-time readout of large-scale unsorted neural ensemble place codes. Cell Rep. 25, 2635-2642. Huang, Q., Jia, J., Han, Q. and Luo, H. (2018).  ... 
doi:10.1101/828467 fatcat:zyuew46vd5ewfemcbzonrl4l7e

Design Methods for Irregular Repeat–Accumulate Codes

A. Roumy, S. Guemghar, G. Caire, S. Verdu
2004 IEEE Transactions on Information Theory  
In this way, the code ensemble optimization can be solved by linear programming.  ...  We optimize the random-like ensemble of irregular repeat-accumulate (IRA) codes for binary-input symmetric channels in the large block-length limit.  ...  The approximations of the DE considered in this work are based on EXIT functions, and track the evolution of the mutual information between the messages output by the bitnodes and the associated code symbols  ... 
doi:10.1109/tit.2004.831778 fatcat:5rxdsqh22zcxzb5pmvp5j6mibu
« Previous Showing results 1 — 15 out of 1,637 results