Publications by Kevin Murphy


All publications have been subject to peer review unless indicated otherwise.


Generative Models of Visually Grounded Imagination
Ramakrishna Vedantam, Ian Fischer, Jonathan Huang, Kevin Murphy
Arxiv, not yet peer reviewed.

PixColor: Pixel Recursive Colorization
Sergio Guadarrama, Ryan Dahl, David Bieber, Mohammad Norouzi, Jonathon Shlens, Kevin Murphy

Attention-based Extraction of Structured Information from Street View Imagery
Zbigniew Wojna, Alex Gorban, Dar-Shyang Lee, Kevin Murphy, Qian Yu, Yeqing Li, Julian Ibarz
Google Research Blog post.

DeepLab: Semantic Image Segmentation with Deep Convolutional Nets, Atrous Convolution, and Fully Connected CRFs
Liang-Chieh Chen, George Papandreou, Iasonas Kokkinos, Kevin Murphy, Alan L. Yuille
IEEE PAMI, 2017.

Deep Probabilistic Programming
Dustin Tran, Matthew D. Hoffman, Rif A. Saurous, Eugene Brevdo, Kevin Murphy, David M. Blei
Open source Tensorflow code.

Deep Variational Information Bottleneck
Alexander A. Alemi, Ian Fischer, Joshua V. Dillon, Kevin Murphy

Semantic Instance Segmentation via Deep Metric Learning
Alireza Fathi, Zbigniew Wojna, Vivek Rathod, Peng Wang, Hyun Oh Song, Sergio Guadarrama, Kevin P. Murphy
Arxiv'17 (not yet peer reviewed).

Context-aware Captions from Context-agnostic Supervision
Ramakrishna Vedantam, Samy Bengio, Kevin Murphy, Devi Parikh, Gal Chechik

Deep Metric Learning via Facility Location
Hyun Oh Song, Stefanie Jegelka, Vivek Rathod, Kevin Murphy

Speed/accuracy trade-offs for modern convolutional object detectors
Jonathan Huang, Vivek Rathod, Chen Sun, Menglong Zhu, Anoop Korattikara, Alireza Fathi, Ian Fischer, Zbigniew Wojna, Yang Song, Sergio Guadarrama, Kevin Murphy
Summary of the open source tensorflow code can be found in this Google Research blog post.
Details on how to run this code on Cloud ML Engine can be found in this Google Cloud ML blog post.
Hacker News.

Towards Accurate Multi-person Pose Estimation in the Wild
George Papandreou, Tyler Zhu, Nori Kanazawa, Alexander Toshev, Jonathan Tompson, Chris Bregler, Kevin Murphy

Optimization of image description metrics using policy gradient methods
Siqi Liu, Zhenhai Zhu, Ning Ye, Sergio Guadarrama, Kevin Murphy


Generation and Comprehension of Unambiguous Object Descriptions
Junhua Mao, Jonathan Huang, Alexander Toshev, Oana Camburu, Alan Yuille, Kevin Murphy
Covered in TechCrunch

Detecting events and key actors in multi-person videos
Vignesh Ramanathan, Jonathan Huang, Sami Abu-El-Haija, Alexander Gorban, Kevin Murphy, Li Fei-Fei.
Covered in TechCrunch

Semantic Image Segmentation with Task-Specific Edge Detection Using CNNs and a Discriminatively Trained Domain Transform
Liang-Chieh Chen, Jonathan T. Barron, George Papandreou, Kevin Murphy, Alan L. Yuille

Efficient inference in occlusion-aware generative models of images
Jonathan Huang, Kevin Murphy
ICLR'16 Workshop.


Im2Calories: towards an automated mobile vision food diary
Austin Myers, Nick Johnston, Vivek Rathod, Anoop Korattikara, Alex Gorban, Nathan Silberman, Sergio Guadarrama, George Papandreou, Jonathan Huang, Kevin Murphy.
Reddit thread. Coverage in popular press: CNET, Fortune, The Guardian, and many many more!

Bayesian Dark Knowledge
Anoop Korattikara, Vivek Rathod, Kevin Murphy, Max Welling

Modeling and Diagnosis of Structural Systems through Sparse Dynamic Graphical Models
Luke Bornn, Charles R Farrar, David Higdon, Kevin Murphy.
J. of Mechanical Systems and Signal Processing, 2015

A Review of Relational Machine Learning for Knowledge Graphs: From Multi-Relational Link Prediction to Automated Knowledge Graph Construction
Maximilian Nickel, Kevin Murphy, Volker Tresp, Evgeniy Gabrilovich
Proc. IEEE, 2015

What's Cookin'? Interpreting Cooking Videos using Text, Speech and Vision
Jon Malmaud, Jonathan Huang, Vivek Rathod, Nicholas Johnston, Andrew Rabinovich, Kevin Murphy

Probabilistic Label Relation Graphs with Ising Models
Nan Ding, Jia Deng, Kevin Murphy, Hartmut Neven
ICCV 2015

TimeMachine: Timeline Generation for Knowledge-Base Entities
Tim Althoff, Xin Luna Dong, Kevin Murphy, Safa Alai, Van Dang, Wei Zhang

Knowledge-Based Trust: Estimating the Trustworthiness of Web Sources
Xin Luna Dong, Evgeniy Gabrilovich, Kevin Murphy, Van Dang, Wilko Horn, Camillo Lugaresi, Shaohua Sun, Wei Zhang
Popular press coverage: CACM, New Scientist, The Independent, Washington Post, Fox News, Salon, other US media outlets

Weakly and semi-supervised learning of a DCNN for semantic image segmentation
George Papandreou, Liang-Chieh Chen, Kevin Murphy, Alan L. Yuille
ICCV 2015

Semantic Image Segmentation with Deep Convolutional Nets and Fully Connected CRFs
Liang-Chieh Chen, George Papandreou, Iasonas Kokkinos, Kevin Murphy, Alan L. Yuille


Probabilistic models for collective entity resoluion between knowledge graphs
Jay Pujara, Kevin Murphy, Luna Dong, Curtis Janssen
Bay Area Machine Learning workshop 2014
(Extended abstract)

Canonicalizing Open Knowledge Bases
Luis Galarraga, Geremy Heitz, Kevin Murphy, Fabian Suchanek
CIKM 2014

Large-Scale Object Classification using Label Relation Graphs
Jia Deng, Nan Ding, Yangqing Jia, Andrea Frome, Kevin Murphy, Samy Bengio, Yuan Li, Hartmut Neven, Hartwig Adam
ECCV 2014 (Best paper award)

Cooking with semantics
Jon Malmaud, Earl Wagner, Nancy Chang, Kevin Murphy.
ACL'14 Semantic Parsing Workshop

Knowledge Vault: A Web-Scale Approach to Probabilistic Knowledge Fusion
Xin Luna Dong, Evgeniy Gabrilovich, Geremy Heitz, Wilko Horn, Ni Lao, Kevin Murphy, Thomas Strohmann, Shaohua Sun, Wei Zhang

From Data Fusion to Knowledge Fusion
Xin Luna Dong, Evgeniy Gabrilovich, Geremy Heitz, Wilko Horn, Kevin Murphy, Shaohua Sun, Wei Zhang

Knowledge Base Completion via Search-Based Question Answering
Robert West, Evgeniy Gabrilovich, Kevin Murphy, Shaohua Sun, Rahul Gupta, Dekang Lin


Extracting Entities and Relations from Web Tables Using a Non-parametric Generative Model
Jon Malmaud, Kevin Murphy
Bay Area Machine Learning workshop (Extended abstract)


Fast Bayesian Inference for Non-Conjugate Gaussian Process Regression
Emtiyaz Khan, Shakir Mohammad, Kevin Murphy
NIPS 2012

Machine learning: a probabilistic perspective
Kevin Murphy
MIT Press 2012
(Not peer reviewed)

Learning to Track and Identify Players from Broadcast Sports Videos
Wei-Lwun Lu, Joanne Ting, Jim Little, Kevin Murphy

Proceedings of the 28th Conference on Uncertainty in Artificial Intelligence
Nando de Freitas and Kevin Murphy (eds)

A Stick-Breaking Likelihood for Categorical Data Analysis with Latent Gaussian Models
M. E. Khan, S. Mohamed, B. Marlin, and K. Murphy
AI/Stats 2012

Efficient Bayesian Inference for Multivariate Probit Models with Sparse Inverse Correlation Matrices
A. Talhouk and A. Doucet and K. Murphy
J. Computational and Graphical Statistics, 21(3), 2012 (official link).


Piecewise Bounds for Estimating Bernoulli-Logistic Latent Gaussian Models
B. Marlin and E. Khan and K. Murphy
ICML 2011
Appendix (truncated Gaussian moments)

Identifying Players in Broadcast Sports Videos using Conditional Random Fields
Wei-Lwun Lu, Jo-Anne Ting, Kevin P. Murphy, and James J. Little
CVPR 2011.
More info on the project
For a more recent version, see the PAMI 2012 paper.

Multiscale Conditional Random Fields for Semi-supervised Labeling and Classification
David Duvenaud, Ben Marlin, Kevin Murphy.
Canadian Conf. on Computer and Robot Vision (CRV) 2011


Variational bounds for mixed-data factor analysis
M. E. Khan, B. Marlin, G. Bouchard, K. Murphy
NIPS 2010.

Pairwise network mechanisms in the host signaling response to coxsackievirus B3 infection
Farshid S. Garmaroudia, David Marchant, Xiaoning Si, Abbas Khalili, Ali Bashashati, Brian W. Wong, Aline Tabet, Raymond T. Ng, Kevin Murphy, Honglin Luo, Kevin A. Janes, Bruce M. McManus.
Proc. Natl. Acad. Sciences, 107(39): 17053-17058, 2010

Computational approaches for RNA energy parameter estimation
M. Andronescu and A. Condon and H. Hoos and K. Murphy and D. Mathews
RNA Journal, 16(12):2304-2118, 2010

Convex Structure Learning in Log-Linear Models: Beyond Pairwise Potentials
M. Schmidt, K. Murphy
AI/Stats 2010.

Time-Bounded Sequential Parameter Optimization
Frank Hutter, Holger Hoos, Kevin Murphy, Kevin Leyton-Brown
Learning and Intelligent Optimization - LION4 2010.
Runner up for best paper award.

Book Review of "Probabilistic graphical models" by Koller and Friedman
Kevin Murphy
Artificial Intelligence Journal, 174(2): 145--146, 2010.
(Invited submission, not peer reviewed)

SNVMix: predicting single nucleotide variants from next generation sequencing of tumors
R. Goya , M. Sun , R. Morin , G. Leung , G. Ha , K. Wieg, , J. Senz , A. Crisan , M. Marra , M. Hirst , D. Huntsman , K. Murphy , S. Aparicio , S. Shah
Bioinformatics, 2010

Using the forest to see the trees: object recognition in context
A. Torralba, K. Murphy, W. Freeman,
Communications of the ACM, Research Highlights, 53(3): 107-114, 2010.
(Invited submission, not peer reviewed)


Accelerating Bayesian Structural Inference for Non-Decomposable Gaussian Graphical Models
Baback Moghaddam, Ben Marlin, Emtiyaz Khan, Kevin Murphy.
NIPS 2009

Causal learning without DAGs
David Duvenaud, Daniel Eaton, Kevin Murphy, Mark Schmidt.
JMLR W&CP 2009.

Group Sparse Priors for Covariance Estimation
Ben Marlin, Mark Schmidt, and Kevin Murphy
UAI 2009

Modeling Discrete Interventional Data using Directed Cyclic Graphical Models
Mark Schmidt, Kevin Murphy
UAI 2009

Sparse Gaussian Graphical Models with Unknown Block Structure
Ben Marlin and Kevin Murphy
ICML 2009

Model based clustering of array CGH data
Sohrab Shah, K-John Cheung, Nathalie Johnson, Randy Gascoyne, Douglas Horsman, Raymond Ng, Kevin Murphy.
Bioinformatics 2009, 25(12):i30-i38.

An Experimental Investigation of Model-Based Parameter Optimisation: SPO and Beyond
Frank Hutter, Kevin Leyton-Brown, Kevin Murphy.
Gecco 2009.

Sequential Model-Based Parameter Optimization: an Experimental Investigation of Automated and Interactive Approaches
F. Hutter, T Bartz-Beielstein, H. Hoos, K. Leyton-Brown, K. Murphy
in Empirical Methods for the Analysis of Optimization Algorithms, 2009.

A Hybrid Conditional Random Field for estimating the underlying ground surface from airborne LiDAR data
Wei-Lwun Lu, Kevin Murphy, James J. Little, Alla Sheffer, Hongbo Fu.
IEEE Trans. on Geoscience and Remote Sensing, 2009, 47(8):2913--2922.

Optimizing Costly Functions with Simple Constraints: A Limited-Memory Projected Quasi-Newton Algorithm
Mark Schmidt, Ewout van den Berg, Michael Friedlander, Kevin Murphy
AI/Stats 2009 (Best paper award)


Structure Learning in Random Fields for Heart Motion Abnormality Detection,
Mark Schmidt, Kevin Murphy, Glenn Fung, Romer Rosales.
CVPR 2008. Appendix. Software.

Genome-wide profiling of follicular lymphoma by array comparative genomic hybridization reveals prognostically significant DNA copy number imbalances
K-J. Cheung, S. Shah, C. Steidl, N. Johnson, T. Relander, A. Telenius, B. Lai, K. Murphy, W. Lam, A. Al-Tourah, J. Connors, R. Ng, R. Gascoyne, D. Horsman.
Blood (J. of the Am. Soc. of Hematology), 2008.

LabelMe: a database and web-based tool for image annotation
Bryan Russell, Antonio Torralba, Kevin Murphy and William Freeman
Intl. J. Computer Vision (special issue on vision and learning), 77(1-3): 157--173, 2008. Software.


Software for graphical models: a review.
Kevin Murphy.
ISBA (Intl. Soc. for Bayesian Analysis) Bulletin, 14(4), pages 13-15, December 2007.
(Invited submission, not peer reviewed.)

Bayesian structure learning using dynamic programming and MCMC
Daniel Eaton and Kevin Murphy
UAI 2007. Software

Modeling changing dependency structure in multivariate time series
Xiang Xuan and Kevin Murphy.
Intl. conf on machine learning (ICML), 2007.

Learning Graphical Model Structure using L1-Regularization Paths
M Schmidt, A Niculescu-Mizil, K Murphy.
AAAI'07. Software

Efficient parameter estimation for RNA secondary structure prediction
M Andronescu, A Condon, H Hoos, D Mathews, K Murphy.
Bioinformatics 2007

Modeling recurrent DNA copy number alterations in array CGH data
S Shah, W Lam, R Ng, K Murphy.
Bioinformatics 2007. Software.

Exact Bayesian structure learning from uncertain interventions
Daniel Eaton and Kevin Murphy.
AI & Statistics, 2007. Software

Sharing visual features for multiclass and multiview object detection
Antonio Torralba, Kevin Murphy and William Freeman
IEEE PAMI, 29(5), May 2007

Figure-ground segmentation using a hierarchical conditional random field
Jordan Reynolds and Kevin Murphy.
Fourth Canadian Conference on Computer and Robot Vision (CRV 2007)

A non-myopic approach to visual search
Julia Vogel and Kevin Murphy.
Fourth Canadian Conference on Computer and Robot Vision (CRV 2007)


Integrating copy number polymorphisms into array CGH analysis using a robust HMM
S Shah, X Xuang, R DeLeeuw, M Khojasteh, W Lam, R Ng, K Murphy
Bioinformatics, 22(14):e431-e439, July 2006. Software.

Accelerated Training of Conditional Random Fields with Stochastic Meta-Descent
S Vishwanathan, N. Schraudolph, M. Schmidt, K. Murphy
ICML'06 (Intl Conf on Machine Learning) Software.


Object detection and localization using local and global features
Kevin Murphy, Antonio Torralba, Daniel Eaton, William Freeman
Appears in Towards Category-Level Object Recognition
LNCS Vol. 4170, 2006, Editors J. Ponce, M. Hebert, C. Schmid, A. Zisserman.
(Invited submission, not peer reviewed)

Shared features for multiclass object detection
Antonio Torralba, Kevin Murphy, William Freeman
Appears in Towards Category-Level Object Recognition
LNCS Vol. 4170, 2006, Editors J. Ponce, M. Hebert, C. Schmid, A. Zisserman.
(Invited submission, not peer reviewed)


Contextual Models for Object Detection using Boosted Random Fields
Antonio Torralba, Kevin Murphy and William Freeman

Sharing features: efficient boosting procedures for multiclass object detection
Antonio Torralba, Kevin Murphy and William Freeman
CVPR'04 (Computer Vision and Pattern Recognition).
[Best poster award]

Representing hierarchical POMDPs as DBNs for multi-scale robot localization
Georgios Theocharous, Kevin Murphy, Leslie Kaelbling
ICRA'04 (Intl. Conf. on Robotics and Automation)


"Using the Forest to See the Trees:A Graphical Model Relating Features, Objects and Scenes"
Kevin Murphy, Antonio Torralba, William Freeman
NIPS'03 (Neural Info. Processing Systems)
More information about this project (including movies) is available.

Context-based vision system for place and object recognition
Antonio Torralba, Kevin Murphy, William Freeman, Mark Rubin
ICCV'03 (Intl. Conf. on Computer Vision)
More information about this project (including movies) is available.


Dynamic Bayesian Networks for Audio-Visual Speech Recognition
A. Nefian, L. Liang, X. Pi, X. Liu and K. Murphy
EURASIP, Journal of Applied Signal Processing, 11:1-15, 2002

A Coupled HMM for Audio-Visual Speech Recognition
A. Nefian, L. Liang, X. Pi, L. Xiaoxiang, C. Mao and K. Murphy
ICASSP '02 (IEEE Int'l Conf on Acoustics, Speech and Signal Proc.), 2:2013--2016.

Learning Markov Processes
Kevin Murphy.
The Encyclopedia of Cognitive Science
L. Nadel et al. (eds), Nature Macmillan, 2002.
(Invited submission, not peer reviewed)


Linear Time Inference in Hierarchical HMMs
Kevin Murphy and Mark Paskin.
NIPS '01 (Neural Info. Proc. Systems).

The Factored Frontier Algorithm for Approximate Inference in DBNs
Kevin Murphy and Yair Weiss.
UAI '01 (Uncertainty in AI).

The Bayes Net Toolbox for Matlab
Kevin Murphy.
Computing Science and Statistics, vol 33, 2001.
(Invited submission, not peer reviewed)
The software is available.

Rao-Blackwellised Particle Filtering for Dynamic Bayesian Networks
Kevin Murphy and Stuart Russell.
Appears in Sequential Monte Carlo Methods in Practice
A. Doucet, N. de Freitas and N.J. Gordon (eds), Springer-Verlag, 2001.


Rao-Blackwellised Particle Filtering for Dynamic Bayesian Networks
Arnaud Doucet, Nando de Freitas, Kevin Murphy and Stuart Russell.
UAI '00 (Uncertainty in AI).


Bayesian Map Learning in Dynamic Environments
Kevin Murphy.
NIPS '99 (Neural Info. Proc. Systems).

Loopy-belief Propagation for Approximate Inference: An Empirical Study
Kevin Murphy, Yair Weiss, and Michael Jordan.
UAI '99 (Uncertainty in AI).

A Variational Approximation for Bayesian Networks with Discrete and Continuous Latent Variables
Kevin Murphy.
UAI '99 (Uncertainty in AI).

A Dynamic Bayesian Network Approach to Figure Tracking Using Learned Dynamic Models
Vladimir Pavlovic, James Rehg, Tat-Jen Cham, and Kevin Murphy.
ICCV '99 (Int'l Conf. on Computer Vision)

Vision-Based Speaker Detection Using Bayesian Networks
James Rehg, Kevin Murphy, and Paul Fieguth.
CVPR '99 (Computer Vision and Pattern Recognition).
(An earlier version of this work appeared in PUI '98 (Perceptual User Interfaces).)


Learning the Structure of Dynamic Probabilistic Networks
Nir Friedman, Kevin Murphy, and Stuart Russell.
UAI '98 (Uncertainty in AI).


Space-efficient Inference in Dynamic Probabilistic Networks
John Binder, Kevin Murphy, and Stuart Russell.
IJCAI '97 (Intl. Joint Conf. on AI).


Automata-Theoretic Models of Mutation and Alignment
David Searls and Kevin Murphy.
ISMB '95 (Intelligent Systems For Molecular Biology).


Talks (not up to date)

Software tookits for machine learning and graphical models
presented at NIPS 2005 Lineal workshop

Exact inference in graphical models
presented at AAAI 2004 tutorial

Approximatte inference in graphical models
presented at AAAI 2004 tutorial

Graphical models and BNT
presented at the Mathworks, May 2003

An introduction to machine learning and graphical models,
presented at the Intel workshop on "Machine learning for the life sciences", Fall 2003

Tutorial on DBNs,
presented at the MIT AI Lab, November 2002

Technical reports/ informal notes

Conjugate Bayesian analysis of the univariate Gaussian: a tutorial
Kevin Murphy, September 2007.

A review of methods for visual object detection
Kevin Murphy, May 2005.

Proposed design for gR, a graphical models toolkit for R
Kevin Murphy, September 2003.

Fitting a constrained conditional linear Gaussian distribution
Kevin Murphy. October 1998, updated January 2003.

Hidden semi-Markov models (segment models)
Kevin Murphy. November 2002.

Dynamic Bayesian Networks (Draft)
To appear in Probabilistic Graphical Models, Michael Jordan.
Kevin Murphy. November 2002.

Tutorial on DBNs (slides)
Kevin Murphy. November 2002.

Representing Hierarchical POMDPs as DBNs, with Applications to Mobile Robot Navigation
Kevin Murphy. November 2002.
This contains some more details than the ICML03 paper above.

Fast manipulation of multi-dimensional arrays in Matlab
Kevin Murphy. September 2002.

Pearl's algorithm for vector Gaussian Bayes Nets
Kevin Murphy. March 2002.

Hierarchical HMMs
Kevin Murphy. November 2001.
This is an extended version of my NIPS'01 paper.
See my thesis (chapter 2) for more information.

Applying the Junction Tree Algorithm to Variable-Length DBNs
Kevin Murphy. October 2001.
See my thesis (chapter 3) for a newer approach.

From Belief Propagation to Expectation Propagation
Kevin Murphy. September 2001.
See my thesis (appendix B) for related material.

Embedded graphical models
Kevin Murphy and Ara Nefian. June 2001.
Intel Research Technical Report.

An introduction to graphical models
Kevin Murphy. May 2001.

Active learning of causal Bayes net structure
Kevin Murphy. March 2001.

Learning Bayes net structure from sparse data sets
Kevin Murphy. February 2001.
See my thesis (appendix C) for more material.

A Survey of POMDP Solution Techniques
Kevin Murphy. September 2000.

Modeling Freeway Traffic using Coupled HMMs
Jaimyoung Kwon and Kevin Murphy. May 2000.

MCMC for Conditionally Linear Gaussian State-Space Models
Kevin Murphy. 2000.

Modelling Gene Expression Data using Dynamic Bayesian Networks
Kevin Murphy and Saira Mian. 1999.

Pearl's algorithm for multiplexer nodes
Kevin Murphy. 1999.

Filtering and Smoothing in Linear Dynamical Systems using the Junction Tree Algorithm
Kevin Murphy. 1998.

Learning Switching Kalman Filter Models
Kevin Murphy. Compaq Cambridge Research Lab Tech Report 98-10, 1998.

Inference and learning in hybrid Bayesian networks
Kevin Murphy. U.C. Berkeley Technical Report CSD-98-990, 1998.

Optimal Alignments in Linear Space using Automaton-Derived Cost Functions
Kevin Murphy. 1996.

Learning Finite Automata,
Kevin Murphy. Santa Fe Institute Technical Report 96-04-017, 1996