Unbiased Learning of Deep Generative Models with Structured Discrete Representations.H. Bendekgey, G. Hope, and E. Sudderth,Neural Information Processing Systems, Dec. 2023.paper
Differentiable and Stable Long-Range Tracking of Multiple Posterior Modes.A. Younis and E. Sudderth,Neural Information Processing Systems, Dec. 2023.paper
A Decoder Suffices for Query-Adaptive Variational Inference.S. Agarwal, G. Hope, A. Younis, and E. Sudderth,Uncertainty in Artificial Intelligence, July 2023.paper · supplement
2022
Prediction-Constrained Markov Models for Medical Time Series with Missing Data and Few Labels.P. Rath, G. Hope, K. Heuton, E. Sudderth, and M. Hughes,Workshop on Learning from Time Series for Health, Conference on Neural Information Processing Systems, Dec. 2022.paper · workshop(spotlight)
Thinned Random Measures for Sparse Graphs with Overlapping Communities.F. Ricci, M. Guindani, and E. Sudderth,Neural Information Processing Systems, Dec. 2022.paper · NeurIPS
Variational Inference for Soil Biogeochemical Models.D. Sujono, H. W. Xie, S. Allison, and E. Sudderth,AI for Science Workshop, International Conference on Machine Learning, July 2022.paper · workshop
Learning Consistent Deep Generative Models from Sparsely Labeled Data.G. Hope, M. Abdrakhmanova, X. Chen, M. Hughes, and E. Sudderth,Symposium on Advances in Approximate Bayesian Inference, Feb. 2022.abstract · AABI
2021
Scalable and Stable Surrogates for Flexible Classifiers with Fairness Constraints.H. Bendekgey and E. Sudderth,Neural Information Processing Systems, Dec. 2021.paper · supplement
Marginalized Stochastic Natural Gradients for Black-Box Variational Inference.G. Ji, D. Sujono, and E. Sudderth,International Conference on Machine Learning, July 2021.paper · supplement
Prediction-Constrained Hidden Markov Models for Semi-Supervised Classification.G. Hope, M. Hughes, F. Doshi-Velez, and E. Sudderth,Time Series Workshop, International Conference on Machine Learning, July 2021.paper · workshop(best poster award)
2020
Learning Consistent Deep Generative Models from Sparse Data via Prediction Constraints.G. Hope, M. Abdrakhmanova, X. Chen, M. Hughes, and E. Sudderth,arXiv:2012.06718, cs.LG, Dec. 2020.arXiv
Clouds of Oriented Gradients for 3D Detection of Objects, Surfaces, and Indoor Scene Layouts.Z. Ren and E. Sudderth,IEEE Trans. on Pattern Analysis & Machine Intelligence, vol. 42(10), Oct. 2020.IEEE · arXiv
Effective Monte Carlo Variational Inference for Binary-Variable Probabilistic Programs.G. Ji and E. Sudderth,International Conference on Probabilistic Programming, Oct. 2020.PROBPROG · abstract · supplement
2019
3D Scene Reconstruction with Multi-layer Depth and Epipolar Transformers.D. Shin, Z. Ren, E. Sudderth, and C. Fowlkes,International Conference on Computer Vision, 2019.project · paper · supplement · arXiv
Variational Training for Large-Scale Noisy-OR Bayesian Networks.G. Ji, D. Cheng, H. Ning, C. Yuan, H. Zhou, L. Xiong, and E. Sudderth,Uncertainty in Artificial Intelligence, July 2019.paper · supplement
A Fusion Approach for Multi-Frame Optical Flow Estimation.Z. Ren, O. Gallo, D. Sun, M-H. Yang, E. Sudderth, and J. Kautz,IEEE Winter Conference on Applications of Computer Vision, Jan. 2019.paper · arXiv
2018
3D Object Detection with Latent Support Surfaces.Z. Ren and E. Sudderth,IEEE Conference on Computer Vision & Pattern Recognition, June 2018.paper
Semi-Supervised Prediction-Constrained Topic Models.M. Hughes, L. Weiner, G. Hope, T. McCoy Jr., R. Perlis, E. Sudderth, and F. Doshi-Velez,Artificial Intelligence & Statistics, Apr. 2018.paper · supplement · arXiv (preliminary version)
2017
Multiscale Semi-Markov Dynamics for Intracortical Brain-Computer Interfaces.D. Milstein, J. Pacheco, L. Hochberg, J. Simeral, B. Jarosiewicz, and E. Sudderth,Neural Information Processing Systems, Dec. 2017.paper · supplement · video
Prediction-Constrained Topic Models for Antidepressant Recommendation.M. Hughes, G. Hope, L. Weiner, T. McCoy Jr., R. Perlis, E. Sudderth, and F. Doshi-Velez,NIPS 2017 Workshop on Machine Learning for Health, December 2017.arXiv
Cascaded Scene Flow Prediction using Semantic Segmentation.Z. Ren, D. Sun, J. Kautz, and E. Sudderth,International Conference on 3D Vision, Oct. 2017.paper · supplement · arXiv
From Patches to Images: A Nonparametric Generative Model.G. Ji, M. Hughes, and E. Sudderth,International Conference on Machine Learning, Aug. 2017.paper · supplement
Refinery: An Open Source Topic Modeling Web Platform.D. Kim, B. Swanson, M. Hughes, and E. Sudderth,Journal of Machine Learning Research, vol. 18, Mar. 2017.paper · jmlr · code · video demo
2016
Fast Learning of Clusters and Topics via Sparse Posteriors.M. Hughes and E. Sudderth,arXiv:1609.07521, stat.ML, September 2016.arXiv
Three-Dimensional Object Detection and Layout Prediction using Clouds of Oriented Gradients.Z. Ren and E. Sudderth,IEEE Conference on Computer Vision & Pattern Recognition, June 2016.paper · supplement
2015
Scalable Adaptation of State Complexity for Nonparametric Hidden Markov Models.M. Hughes, W. Stephenson, and E. Sudderth,Neural Information Processing Systems, Dec. 2015.paper · supplement
Proteins, Particles, and Pseudo-Max-Marginals: A Submodular Approach.J. Pacheco and E. Sudderth,International Conference on Machine Learning, July 2015.paper
Layered RGBD Scene Flow Estimation.D. Sun, E. Sudderth, and H. Pfister,IEEE Conference on Computer Vision & Pattern Recognition, June 2015.paper
Reliable and Scalable Variational Inference for the Hierarchical Dirichlet Process.M. Hughes, D. Kim, and E. Sudderth,Artificial Intelligence & Statistics, May 2015.paper
A Spectral Clustering Search Algorithm for Predicting Shallow Landslide Size and Location.D. Bellugi, D. Milledge, W. Dietrich, J. McKean, J. Perron, E. Sudderth, and B. Kazian,JGR: Earth Surface, vol. 120, 2015.paper · JGR · AGU research spotlight
2014
Joint Modeling of Multiple Time Series via the Beta Process with Application to Motion Capture Segmentation.E. Fox, M. Hughes, E. Sudderth, and M. Jordan,Annals of Applied Statistics, vol. 8(3), 2014.paper
Nonparametric Clustering with Distance Dependent Hierarchies.S. Ghosh, M. Raptis, L. Sigal, and E. Sudderth,Uncertainty in Artificial Intelligence, July 2014.paper
Preserving Modes and Messages via Diverse Particle Selection.J. Pacheco, S. Zuffi, M. Black, and E. Sudderth,International Conference on Machine Learning, June 2014.paper
Quantifying Aphid Behavioral Responses to Environmental Change.E. A. Sudderth and E. B. Sudderth,Entomologia Experimentalis et Applicata, vol. 150, 2014.paper
2013
Memoized Online Variational Inference for Dirichlet Process Mixture ModelsM. Hughes and E. SudderthNeural Information Processing Systems, Dec. 2013.paper
Efficient Online Inference for Bayesian Nonparametric Relational ModelsD. Kim, P. Gopalan, D. Blei, and E. SudderthNeural Information Processing Systems, Dec. 2013.paper
A Fully-Connected Layered Model of Foreground and Background FlowD. Sun, J. Wulff, E. Sudderth, H. Pfister, and M. BlackIEEE Conference on Computer Vision & Pattern Recognition, June 2013.paper
NET-VISA: Network Processing Vertically Integrated Seismic AnalysisN. Arora, S. Russell, and E. SudderthBulletin of the Seismological Society of America, vol. 103(2a), Apr. 2013.paper
2012
Truly Nonparametric Online Variational Inference for Hierarchical Dirichlet ProcessesM. Bryant and E. SudderthNeural Information Processing Systems, Dec. 2012.paper
From Deformations to Parts: Motion-based Segmentation of 3D ObjectsS. Ghosh, E. Sudderth, M. Loper, and M. BlackNeural Information Processing Systems, Dec. 2012.paper
Effective Split-Merge Monte Carlo Methods for Nonparametric Models of Sequential DataM. Hughes, E. Fox, and E. SudderthNeural Information Processing Systems, Dec. 2012.paper
Minimization of Continuous Bethe Approximations: A Positive VariationJ. Pacheco and E. SudderthNeural Information Processing Systems, Dec. 2012.paper
Improved Variational Inference for Tracking in ClutterJ. Pacheco and E. SudderthIEEE Statistical Signal Processing Workshop, Aug. 2012.paper
The Nonparametric Metadata Dependent Relational ModelD. Kim, M. Hughes, and E. SudderthInternational Conference on Machine Learning, June 2012.paper · supplement
Layered Segmentation and Optical Flow Estimation Over TimeD. Sun, E. Sudderth, and M. BlackIEEE Conference on Computer Vision & Pattern Recognition, June 2012.paper
Nonparametric Learning for Layered Segmentation of Natural ImagesS. Ghosh and E. SudderthIEEE Conference on Computer Vision & Pattern Recognition, June 2012.paper
Nonparametric Discovery of Activity Patterns from Video CollectionsM. Hughes and E. SudderthCVPR Workshop on Perceptual Organization in Computer Vision.paper
2011
The Doubly Correlated Nonparametric Topic ModelD. Kim and E. SudderthNeural Information Processing Systems, Dec. 2011.paper
Spatial Distance Dependent Chinese Restaurant Processes for Image SegmentationS. Ghosh, A. Ungureanu, E. Sudderth, and D. BleiNeural Information Processing Systems, Dec. 2011.paper
A Sticky HDP-HMM with Application to Speaker DiarizationE. Fox, E. Sudderth, M. Jordan, and A. WillskyAnnals of Applied Statistics, vol. 5(2A), 2011.paper · arXiv
Bayesian Nonparametric Inference of Switching Dynamic Linear ModelsE. Fox, E. Sudderth, M. Jordan, and A. WillskyIEEE Transactions on Signal Processing, vol. 59(4), Apr. 2011.paper
Global Seismic Monitoring: A Bayesian ApproachN. Arora, S. Russell, P. Kidwell, and E. SudderthAAAI Conference on Artificial Intelligence, Nectar track, 2011.paper
2010
Layered Image Motion with Explicit Occlusions, Temporal Consistency, and Depth OrderingD. Sun, E. Sudderth, and M. BlackNeural Information Processing Systems, Dec. 2010.paper
Global Seismic Monitoring as Probabilistic InferenceN. Arora, S. Russell, P. Kidwell, and E. SudderthNeural Information Processing Systems, Dec. 2010.paper
Bayesian Nonparametric Learning of Markov Switching ProcessesE. Fox, E. Sudderth, M. Jordan, and A. WillskyIEEE Signal Processing Magazine, vol. 27(6), Nov. 2010.paper
Nonparametric Belief PropagationE. Sudderth, A. Ihler, M. Isard, W. Freeman, and A. WillskyCommunications of the ACM, vol. 53(10), Oct. 2010.paper
Gibbs Sampling in Open-Universe Stochastic LanguagesN. Arora, R. de Salvo Braz, E. Sudderth, and S. RussellUncertainty in Artificial Intelligence, July 2010.paper
2009
Sharing Features among Dynamical Systems with Beta ProcessesE. Fox, E. Sudderth, M. Jordan, and A. WillskyNeural Information Processing Systems, Dec. 2009.paper
Nonparametric Belief Propagation for Distributed Tracking of Robot Networks with Noisy Inter-Distance MeasurementsJ. Schiff, E. Sudderth, and K. GoldbergIEEE International Conference on Intelligent Robots and Systems, Oct. 2009.paper
Nonparametric Bayesian Identification of Jump Systems with Sparse DependenciesE. Fox, E. Sudderth, M. Jordan, and A. WillskyIFAC Symposium on System Identification, July 2009.paper
2008
Shared Segmentation of Natural Scenes Using Dependent Pitman-Yor ProcessesE. Sudderth and M. JordanNeural Information Processing Systems, Dec. 2008.paper · slides
Nonparametric Bayesian Learning of Switching Linear Dynamical SystemsE. Fox, E. Sudderth, M. Jordan, and A. WillskyNeural Information Processing Systems, Dec. 2008.paper
An HDP-HMM for Systems with State PersistenceE. Fox, E. Sudderth, M. Jordan, and A. WillskyInternational Conference on Machine Learning, July 2008.paper
Describing Visual Scenes Using Transformed Objects and PartsE. Sudderth, A. Torralba, W. Freeman, and A. WillskyInternational Journal of Computer Vision, vol. 77, Mar. 2008.paper
Signal and Image Processing with Belief PropagationE. Sudderth and W. FreemanIEEE Signal Processing Magazine, DSP Applications Column, Mar. 2008.paper
2007
Loop Series and Bethe Variational Bounds in Attractive Graphical ModelsE. Sudderth, M. Wainwright, and A. WillskyNeural Information Processing Systems, Dec. 2007.paper
Learning Multiscale Representations of Natural Scenes Using Dirichlet ProcessesJ. Kivinen, E. Sudderth, and M. JordanIEEE International Conference on Computer Vision, Oct. 2007.paper
Image Denoising with Nonparametric Hidden Markov TreesJ. Kivinen, E. Sudderth, and M. JordanIEEE International Conference on Image Processing, Sep. 2007.paper
Hierarchical Dirichlet Processes for Tracking Maneuvering TargetsE. Fox, E. Sudderth, and A. WillskyInternational Conference on Information Fusion, July 2007.paper
2006
Depth from Familiar Objects: A Hierarchical Model for 3D ScenesE. Sudderth, A. Torralba, W. Freeman, and A. WillskyIEEE Conference on Computer Vision & Pattern Recognition, June 2006.paper
2005
Describing Visual Scenes using Transformed Dirichlet ProcessesE. Sudderth, A. Torralba, W. Freeman, and A. WillskyNeural Information Processing Systems, Dec. 2005.paper
Learning Hierarchical Models of Scenes, Objects, and PartsE. Sudderth, A. Torralba, W. Freeman, and A. WillskyInternational Conference on Computer Vision, Oct. 2005. paper
2004
Distributed Occlusion Reasoning for Tracking with Nonparametric Belief PropagationE. Sudderth, M. Mandel, W. Freeman, and A. WillskyNeural Information Processing Systems, Dec. 2004.paper
Embedded Trees: Estimation of Gaussian Processes on Graphs with CyclesE. Sudderth, M. Wainwright, and A. WillskyIEEE Transactions on Signal Processing, vol. 52(11), Nov. 2004.paper
Visual Hand Tracking Using Nonparametric Belief PropagationE. Sudderth, M. Mandel, W. Freeman, and A. WillskyWorkshop on Generative Model Based Vision, CVPR, June 2004.paper
2003
Efficient Multiscale Sampling from Products of Gaussian MixturesA. Ihler, E. Sudderth, W. Freeman, and A. WillskyNeural Information Processing Systems, Dec. 2003.paper
Nonparametric Belief PropagationE. Sudderth, A. Ihler, W. Freeman, and A. WillskyIEEE Conference on Computer Vision & Pattern Recognition, June 2003.paper
2002 & earlier
Statistical and Information-Theoretic Methods for Self-Organization and Fusion of Multimodal, Networked SensorsJ. Fisher III, M. Wainwright, E. Sudderth, and A. WillskyInt. Journal of High Performance Computing Applications, vol. 16(3), Fall 2002.paper
Projection Algebra Analysis of Error-Correcting CodesJ. Yedidia, E. Sudderth, and J-P. BouchaudAllerton Conference on Communication, Control, and Computing, Oct. 2001.paper
Tree-Based Modeling and Estimation of Gaussian Processes on Graphs with CyclesM. Wainwright, E. Sudderth, and A. WillskyNeural Information Processing Systems, Dec. 2000.paper
Theses
Graphical Models for Visual Object Recognition and TrackingE. B. SudderthDoctoral Thesis, Massachusetts Institute of Technology, May 2006.thesis
Embedded Trees: Estimation of Gaussian Processes on Graph with CyclesE. B. SudderthMasters Thesis, Massachusetts Institute of Technology, Feb. 2002.thesis
A Kinematic Model Compiler for the Estimation of Articulated Motion from Video SequencesE. B. SudderthSenior Honors Thesis, University of California at San Diego, April 1999.thesis
Selected Talks
Diverse Particle Selection for High-Dimensional Inference in Graphical ModelsE. Sudderth, J. Pacheco, S. Zuffi, and M. BlackSouthern California Machine Learning Symposium, October 2017.slides
Reliable Variational Learning for Hierarchical Dirichlet ProcessesE. Sudderth, M. Hughes, and D. KimNIPS Workshop on Advances in Variational Inference, December 2014.slides
Reliable Variational Learning for Hierarchical Dirichlet ProcessesE. Sudderth, M. Hughes, D. Kim, P. Gopalan, and D. BleiInternational Society for Bayesian Analysis World Meeting, July 2014.slides
Toward Reliable Bayesian Nonparametric LearningE. Sudderth, D. Wei, M. Bryant, M. Hughes, and E. FoxNIPS Workshop on Bayesian Nonparametric Models for Reliable Planning and Decision-Making Under Uncertainty, Dec. 2012.slides
Spatial Bayesian Nonparametrics for Natural Image SegmentationE. Sudderth, M. Jordan, and S. GhoshNIPS Workshop on Bayesian Nonparametrics: Hope or Hype? Dec. 2011.slides
Representation in Low-Level Visual LearningE. SudderthNSF Workshop on Frontiers in Computer Vision, Aug. 2011.slides
Visual Learning via Topics, Transformations, and TreesE. SudderthNIPS Workshop on Transfer Learning by Learning Rich Generative Models, Dec. 2010.slides
Loop Series and Bethe Variational Bounds in Attractive Graphical ModelsE. Sudderth, M. Wainwright, and A. WillskyAllerton Conference on Communication, Control, and Computing, Oct. 2007.slides