Nebojsa Jojic

A comparison of algorithms for inference and learning in probabilistic graphical models (2009)

Brendan J. Frey, Senior Member, Nebojsa Jojic

Abstract—Research into methods for reasoning under uncertainty is currently one of the most exciting areas of artificial intelligence, largely because it has recently become possible to record,...

POPULATION SEQUENCING USING SHORT READS: HIV AS A CASE STUDY (2009)

Vladimir Jojic, Tomer Hertz, Nebojsa Jojic

Despite many drawbacks, traditional sequencing technologies have proven to be invaluable in modern medical research, even when the targeted genomes are highly variable. While it is often known in...

Fast Transformation Invariant Component Analysis (2008)

Anitha Kannan, Nebojsa Jojic, Brendan J. Frey

Dimensionality reduction techniques such as principal component analysis and factor analysis are used to discover a linear mapping between high dimensional data samples and points in a lower...

1000 Transformation-Invariant Clustering and Dimensionality Reduction Using EM Abstract Submitted to IEEE Transactions on Pattern Analysis and Machine Intelligence, Nov. 2000. (2008)

Brendan J. Frey, Nebojsa Jojic

Clustering and dimensionality reduction are simple, effective ways to derive useful representations of data, such as images. These procedures often are used as preprocessing steps for more...

Epitomic Analysis of Human Motion (2008)

Wooyoung Kim, Nebojsa Jojic, James M. Rehg

The image epitome was introduced by Jojic, Frey and Kannan in [1] as a novel generative image model. The epitome representation of a source image is much smaller than the source but retains it....

GENERATIVE MODELS FOR COMPUTER VISION (2008)

Nebojsa Jojic, Thomas S. Huang, Brendan J. Frey

In order to build robust computer vision algorithms, scene models are necessary that are capable of capturing various aspects of the data at the same time. These models should be fairly simple, but...

A Generative Model of Dense Optical Flow in Layers (2008)

Anitha Kannan, Brendan Frey, Nebojsa Jojic

Abstract. We introduce a generative model of dense flow fields within a layered representation of 3-dimensional scenes. Using probabilistic inference and learning techniques (namely, variational...

See a PPT file with videos at www.research.microsoft.com/users/jojic/FlexiblesSprites.htm Learning Flexible Sprites in Video Layers (2008)

Nebojsa Jojic, Brendan J. Frey

We propose a technique for automatically learning layers of “flexible sprites ” – probabilistic 2dimensional appearance maps and masks of moving, occluding objects. The model explains each...

for video demos Video epitomes (2008)

Vincent Cheung, Brendan J. Frey, Nebojsa Jojic

Recently, “epitomes ” were introduced as patch-based probability models that are learned by compiling together a large number of examples of patches from input images. In this paper, we describe...

Automatic object discovery (2008)

Nebojsa Jojic, John Winn, Larry Zitnick

local minima through hierarchical model selection:

The (2008)

Manuel Reyes-gomez, Nebojsa Jojic

single-channel unsupervised source separation of speech mixtures:

Q-Clustering (2008)

Mukund Narasimhan, Nebojsa Jojic, Jeff Bilmes

We show that Queyranne’s algorithm for minimizing symmetric submodular functions can be used for clustering with a variety of different objective functions. Two specific criteria that we consider...

Q-Clustering (2008)

Mukund Narasimhan, Nebojsa Jojic, Jeff Bilmes

We show that Queyranne’s algorithm for minimizing symmetric submodular functions can be used for clustering with a variety of different objective functions. Two specific criteria that we consider...

Abstract (2008)

Brendan J. Frey, Nebojsa Jojic

www.ifp.uiuc.edu/∼jojic In previous work on “transformed mixtures of Gaussians ” and “transformed hidden Markov models”, we showed how the EM algorithm in a discrete latent variable model...

Using “epitomes ” to model genetic diversity: Rational design of HIV vaccine cocktails (2008)

Nebojsa Jojic, Vladimir Jojic, Brendan Frey, Chris Meek, David Heckerman

We introduce a new model of genetic diversity which summarizes a large input dataset into an epitome, a short sequence or a small set of short sequences of probability distributions capturing many...

Estimating smooth deformation models of substance and noise (2008)

Nebojsa Jojic, Brendan J. Frey, Patrice Simard, David Heckerman

By representing image prototypes, or “substance”, by linear subspaces spanned by deformation fields derived from low-frequecy wavelets, impressive invariance to distortion can be built into...

Computer Modeling, Analysis and Synthesis of Dressed Humans (2008)

Nebojsa Jojic, Jin Gu, Helen C. Shen, Thomas S. Huang

1 In this paper we present computer vision techniques for building dressed human models using images. In the first part of the paper we develop an algorithm for 3-D body reconstruction and texture...

Audio-Visual Graphical Models for Speech Processing (2008)

John Hershey, Hagai Attias, Nebojsa Jojic, Trausti Kristjansson

Perceiving sounds in a noisy environment is a challenging problem. Visual lip-reading can provide relevant information but is also challenging because lips are moving and a tracker must deal with a...

1 (2007)

Nebojsa Jojic, Brendan J. Frey, Patrice Simard, David Heckerman

Learning mixtures of smooth, nonuniform deformation models for probabilistic image matching

-D Reconstruction of Multipart Self-Occluding Objects (2007)

Nebojsa Jojic, Jin Gu, Helen C. Shen, Thomas Huang

. In this paper we present a method for reconstruction of multipart objects from several arbitrary views using deformable superquadrics as the models of the object's parts. Two visual cues are...

Computer Modeling, Analysis and Synthesis of Dressed Humans (2007)

Nebojsa Jojic, Jin Gu, Ivan Mak, Helen C. Shen, Thomas S. Huang

1 In this paper we present a method for 3-D reconstruction of human bodies with application in CAD systems for garment design. The reconstruction scheme uses image information from several arbitrary...

1 (2007)

Brendan J. Frey, Anitha Kannan, Nebojsa Jojic

Factor analysis and principal components analysis can be used to model linear relationships between observed variables and linearly map high-dimensional data to a lower-dimensional hidden space. In...

Coping with Viral Diversity in HIV Vaccine Design (2007)

David C. Nickle, Morgane Rolland, Mark A. Jensen, Wenjie Deng, Mark Seligman, ...

The ability of human immunodeficiency virus type 1 (HIV-1) to develop high levels of genetic diversity, and thereby acquire mutations to escape immune pressures, contributes to the difficulties in...

Learning MHC I--peptide binding (2006)

Jojic, Nebojsa, Reyes-Gomez, Manuel, Heckerman, David, Kadie, Carl, Schueler-Furman, Ora

Motivation and results: Motivated by the ability of a simple threading approach to predict MHC I—peptide binding, we developed a new and improved structure-based model for which parameters can be...

Deformable spectrograms (2005)

Manuel Reyes-gomez, Nebojsa Jojic

Speech and other natural sounds show high temporal correlation and smooth spectral evolution punctuated by a few, irregular and abrupt changes. In a conventional Hidden Markov Model (HMM), such...

A comparison of algorithms for inference and learning in probabilistic graphical models (2005)

Brendan J. Frey, Nebojsa Jojic

Computer vision is currently one of the most exciting areas of artificial intelligence re-search, largely because it has recently become possible to record, store and process large amounts of visual...

Generative Model for Layers of Appearance and Deformation (2005)

Anitha Kannan, Nebojsa Jojic, Brendan J. Frey

We are interested in learning generative models of objects that can be used in wide range of tasks such as video summarization, image segmentation and frame interpolation. Learning object-based...

Q-Clustering (2005)

Mukund Narasimhan, Nebojsa Jojic, Jeff Bilmes

We show that Queyranne's algorithm for minimizing symmetric submodular functions can be used for clustering with a variety of different objective functions. Two specific criteria that we...

A comparison of algorithms for inference and learning in probabilistic graphical models (2005)

Brendan J. Frey, Nebojsa Jojic

Research into methods for reasoning under uncertainty is currently one of the most excit-ing areas of artificial intelligence, largely because it has recently become possible to record, store and...

Video Epitomes (2005)

Vincent Cheung, Brendan J. Frey, Nebojsa Jojic

Recently, “epitomes ” were introduced as patch-based probability models that are learned by compiling together a large number of examples of patches from input images. In this paper, we describe...

Multiband Audio Modeling for Single-Channel Acoustic Source Separation (2004)

Nebojsa Jojic

Detailed hidden Markov models (HMMs) that capture the constraints implicit in a particular sound can be used to estimate obscured or corrupted portions from partial observations, the situation...

Probability models for high dynamic range imaging (2004)

Chris Pal, Rick Szeliski, Matthew Uyttendaele, Nebojsa Jojic

Methods for expanding the dynamic range of digital photographs by combining images taken at different exposures have recently received a lot of attention. Current techniques assume that the...

Probabilistic index maps for modeling natural signals,” UAI (2004)

Nebojsa Jojic, Redmond Wa

One of the major problems in modeling natural signals is that signals with very similar structure may locally have completely different measurements, e.g., images taken under different illumination...

Capturing image structure with probabilistic index maps (2004)

Nebojsa Jojic, Yaron Caspi

One of the major problems in modeling images for vision tasks is that images with very similar structure may locally have completely different appearance, e.g., images taken under different...

Probabilistic index maps for modeling natural signals,” UAI (2004)

Nebojsa Jojic, Redmond Wa

One of the major problems in modeling natural signals is that signals with very similar structure may locally have completely different measurements, e.g., images taken under different illumination...

Efficient approximations for learning phylogenetic HMM models from data (2004)

Jojic, Vladimir, Jojic, Nebojsa, Meek, Chris, Geiger, Dan, Siepel, Adam, Haussler, David, ...

Motivation: We consider models useful for learning an evolutionary or phylogenetic tree from data consisting of DNA sequences corresponding to the leaves of the tree. In particular, we consider a...

Joint Design of Data Analysis Algorithms and User Interface for Video Applications (2003)

Jojic, Nebojsa, Basu, Sumit, Petrovic, Nemanja, Frey, Brendan, Huang, Thomas

The graphical modeling paradigm provides a way of representing data through hidden causes of variability which can be estimated from the data in an unsupervised manner. Recently, a lot of research...

Advances in algorithms for inference and learning in complex probability models (2003)

Brendan J. Frey, Nebojsa Jojic

Computer vision is currently one of the most exciting areas of artificial intelligence research, largely because it has recently become possible to record, store and process large amounts of visual...

Advances in algorithms for inference and learning in complex probability models (2003)

Brendan J. Frey, Nebojsa Jojic

Computer vision is currently one of the most exciting areas of artificial intelligence research, largely because it has recently become possible to record, store and process large amounts of visual...

Joint Design of Data Analysis Algorithms and User (2003)

Interface For Video, Nebojsa Jojic, Sumit Basu, Thomas Huang, Nemanja Petrovic, Brendan Frey

The graphical modeling paradigm provides a way of representing data through hidden causes of variability which can be estimated from the data in an unsupervised manner. Recently a lot of research has...

A Graphical Model for Audiovisual Object Tracking (2003)

Matthew J. Beal, Nebojsa Jojic, Ieee Computer Society, Hagai Attias

We present a new approach to modeling and processing multimedia data. This approach is based on graphical models that combine audio and video variables. We demonstrate it by developing a new...

Fast transformation-invariant factor analysis (2003)

Anitha Kannan, Nebojsa Jojic, Brendan Frey

Dimensionality reduction techniques such as principal component analysis and factor analysis are used to discover a linear mapping between high dimensional data samples and points in a lower...

Fast transformationinvariant component analysis (2003)

Anitha Kannan, Nebojsa Jojic, Brendan J. Frey

Dimensionality reduction techniques such as principal component analysis and factor analysis are used to discover a linear mapping between high dimensional data samples and points in a lower...

Transformation-invariant clustering using the EM algorithm (2003)

Brendan J. Frey, Nebojsa Jojic, Ieee Computer Society, Ieee Computer Society

Abstract—Clustering is a simple, effective way to derive useful representations of data, such as images and videos. Clustering explains the input as one of several prototypes, plus noise. In...

Advances in algorithms for inference and learning in complex probability models (2003)

Brendan J. Frey, Nebojsa Jojic

Computer vision is currently one of the most exciting areas of artificial intelligence research, largely because it has recently become possible to record, store and process large amounts of visual...

Epitomic analysis of appearance and shape (2003)

Nebojsa Jojic, Brendan J. Frey, Anitha Kannan

www.research.microsoft.com/∼jojic www.psi.toronto.edu www.psi.toronto.edu We present novel simple appearance and shape models that we call epitomes. The epitome of an image is its miniature,...

A graphical model for audiovisual object tracking (2003)

Matthew J. Beal, Nebojsa Jojic, Ieee Computer Society, Hagai Attias

Abstract—We present a new approach to modeling and processing multimedia data. This approach is based on graphical models that combine audio and video variables. We demonstrate it by developing a...

Learning appearance and transparency manifolds of occluded objects in layers (2003)

Brendan J. Frey, Nebojsa Jojic, Anitha Kannan

Videos and software available at www.psi.toronto.edu/layers.html By mapping a set of input images to points in a lowdimensional manifold or subspace, it is possible to efficiently account for a small...

A self-calibrating algorithm for speaker tracking based on audio-visual statistical models (2002)

Matthew J. Beal, Nebojsa Jojic, Hagai Attias

We present a self-calibrating algorithm for audio-visual tracking using two microphones and a camera. The algorithm uses a parametrized statistical model which combines simple models of video and...

Audio-video sensor fusion with probabilistic graphical models (2002)

Matthew J. Beal, Hagai Attias, Nebojsa Jojic

Abstract. We present a new approach to modeling and processing multimedia data. This approach is based on graphical models that combine audio and video variables. We demonstrate it by developing a...

A self-calibrating algorithm for speaker tracking based on audio-visual statistical models (2002)

Matthew J. Beal, Nebojsa Jojic, Hagai Attias

We present a self-calibrating algorithm for audio-visual tracking using two microphones and a camera. The algorithm uses a parametrized statistical model which combines simple models of video and...

Fast, large-scale transformation-invariant clustering (2002)

Brendan J. Frey, Nebojsa Jojic

www.ifp.uiuc.edu/#jojic In previous work on "transformed mixtures of Gaussians " and "transformed hidden Markov models", we showed how the EM algorithm in a...

Audio-video sensor fusion with probabilistic graphical models (2002)

Matthew J. Beal, Hagai Attias, Nebojsa Jojic

Abstract. We present a new approach to modeling and processing multimedia data. This approach is based on graphical models that combine audio and video variables. We demonstrate it by developing a...

Learning Montages of Transformed Latent (2002)

Images As Representations, Brendan J. Frey, Nebojsa Jojic

This paper introduces a novel probabilistic model for representing objects that change in appearance as a result of changes in pose, due to small deformations of their sub-parts and the relative...

Audio-video sensor fusion with probabilistic graphical models (2002)

Matthew J. Beal, Hagai Attias, Nebojsa Jojic

Abstract. We present a new approach to modeling and processing multimedia data. This approach is based on graphical models that combine audio and video variables. We demonstrate it by developing a...

Learning flexible sprites in video layers (2001)

Nebojsa Jojic, Brendan J. Frey

We propose a technique for automatically learning layers of \ exible sprites " { probabilistic 2-dimensional appearance maps and masks of moving, occluding objects. The model explains each...

Separating appearance from deformation (2001)

Nebojsa Jojic, Patrice Simard, Brendan J. Frey, David Heckerman

By representing images and image prototypes by linear subspaces spanned by “tangent vectors ” (derivatives of an image with respect to translation, rotation, etc.), impressive invariance to known...

Separating appearance from deformation (2001)

Nebojsa Jojic, Patrice Simard, Brendan J. Frey, David Heckerman

By representing images and image prototypes by linear subspaces spanned by “tangent vectors ” (derivatives of an image with respect to translation, rotation, etc.), impressive invariance to known...

Transformation-invariant clustering and dimensionality reduction using em (2000)

Brendan J. Frey, Nebojsa Jojic

Clustering and dimensionality reduction are simple, effective ways to derive useful representations of data, such as images. These procedures often are used as preprocessing steps for more...

Topographic transformation as a discrete latent variable (2000)

Nebojsa Jojic, Brendan J. Frey

www.ifp.uiuc.edu/jojic www.cs.uwaterloo.ca/frey Invariance to topographic transformations such as translation and shearing in an image has been successfully incorporated into feedforward mechanisms,...

Transformed Hidden Markov Models: Estimating Mixture Models of Images and Inferring Spatial Transformations in Video Sequences (2000)

Nebojsa Jojic, Nemanja Petrovic, Brendan J. Frey, Thomas S. Huang

Submitted to the IEEE Conference on Computer Vision and Pattern Recognition, 2000. In this paper we describe a novel generative model for video analysis called the transformed hidden Markov model...

Detection and Estimation of Pointing Gestures in Dense Disparity Maps (2000)

Nebojsa Jojic, Barry Brumitt, Brian Meyers, Steve Harris, Thomas Huang

In this paper we describe a real-time system for detecting pointing gestures and estimating the direction of pointing using stereo cameras. Previously, similar systems where implemented using...

Tracking Articulated Structures inStereo Image Sequences (1999)

Nebojsa Jojic, Thomas S. Huang

In this paper, we present an algorithm for real time 3-D tracking of articulated structures in stereo image sequences. These sequences can be captured by an inexpensive commercially available system...

Transformed Component Analysis: Joint Estimation of Spatial Transformations and Image Components (1999)

Brendan Frey Nebojsa, Brendan J. Frey, Nebojsa Jojic

Presented at the IEEE International Conference on Computer Vision, Kerkyra, Greece, Sept. 20-25, 1999. A simple, eective way to model images is to represent each input pattern by a linear combination...

Estimating Mixture Models of Images and Inferring Spatial Transformations Using the EM Algorithm (1999)

Brendan Frey Nebojsa, Brendan J. Frey, Nebojsa Jojic

Presented at the IEEE Conference on Computer Vision and Pattern Recognition, Ft. Collins, CO, June, 1999. Mixture modeling and clustering algorithms are effective, simple ways to represent images...

Transformed Component Analysis: Joint Estimation of Spatial Transformations and Image Components (1999)

Brendan Frey Nebojsa, Brendan J. Frey, Nebojsa Jojic

Submitted to the IEEE International Conference on Computer Vision, Kerkyra, Greece, Sept. 20-25, 1999. A simple, eective way to model images is to represent each input pattern by a linear combination...

Estimating Mixture Models of Images and Inferring Spatial Transformations Using the EM Algorithm (1999)

Brendan Frey Nebojsa, Brendan J. Frey, Nebojsa Jojic

Mixture modeling and clustering algorithms are effective, simple ways to represent images using a set of data centers. However, in situations where the images include background clutter and...

3-D reconstruction of multipart self-occluding objects (1998)

Jojic, Nebojsa, Gu, Jin, Shen, Helen C., Huang, Thomas

In this paper we present a method for reconstruction of multipart objects from several arbitrary views using deformable superquadrics as the models of the object's parts. Two visual cues are used:...

Estimating Cloth Draping Parameters from Range Data (1997)

Nebojsa Jojic, Thomas S. Huang

In this paper we present a computer vision algorithm for estimating cloth parameters for modeling. In an analysis-by-synthesis manner, the algorithm compares the drape of the model with the range...

A Generalized Feature Extractor using Expansion Matching and the Karhunen-Loeve Transform (1996)

Dibyendu Nandy, Zhiqian Wang, Jezekiel Ben-arie, K. Raghunath Rao, Nebojsa Jojic

This paper presents a novel generalized feature extraction method based on the Expansion Matching (EXM) method and the KarhunenLoeve (KL) transform. This yields an efficient method to locate a large...

Interactive montages of sprites for indexing and summarizing security video (1192)

Chris Pal, Nebojsa Jojic

In this video we present a new model of interaction for indexing and visualizing video in the context of security applications. We wish to index security video that contains relatively rare but...

Coping with Viral Diversity in HIV Vaccine Design

Nickle, David C, Rolland, Morgane, Jensen, Mark A, Pond, Sergei L. Kosakovsky, Deng, Wenjie, Seligman, Mark, ...

The ability of human immunodeficiency virus type 1 (HIV-1) to develop high levels of genetic diversity, and thereby acquire mutations to escape immune pressures, contributes to the difficulties in...

On Analysis of Cloth Drape Range Data

Nebojsa Jojic, Thomas S. Huang

. In this paper we present an algorithm for analysis of the range data of cloth drapes. The goal of the analysis is estimation of parameters for modeling and the geometry of the underlying object. In...