Erik B. Sudderth, Statistical Computation & Perception

I am a Professor of Computer Science and Statistics at the University of California, Irvine. My Learning, Inference, & Vision Group develops statistical methods for scalable machine learning, with applications in artificial intelligence, computer vision, and the natural and social sciences. My research affiliations at UC Irvine include:

• Director of the UCI Center for Machine Learning and Intelligent Systems. See our seminar series.
• Director of the HPI Research Center in Machine Learning and Data Science at UC Irvine.
• UCI Computational Vision Group.
• UCI Algorithms, Combinatorics, and Optimization Center. See their seminar series.
• UCI Steckler Center for Responsible, Ethical, and Accessible Technology (CREATE).
• UCI Initiative in AI, Law, & Society.
• UCI Center for Theoretical Behavioral Sciences.
• UCI Data Science Initiative. Also see the statistics seminar series.

For a tutorial introduction to probabilistic modeling and approximate inference, see the background chapter of my doctoral thesis, advised by Professors Alan Willsky and William Freeman at MIT EECS. My postdoctoral research at Berkeley EECS, advised by Professors Michael Jordan and Stuart Russell, focused on Bayesian nonparametric models (see my CVPR tutorial).

For more information: bio · curriculum vitæ · research projects & code · publications & lectures

Research Highlights

• New variational inference algorithms enable uncertainty-aware answers for post-training queries of deep generative models. Methods for image inpainting and superresolution with denoising diffusion models appear at the 2025 ICLR Workshop on Frontiers in Probabilistic Inference, generalizing prior work on hierarchical variational autoencoders (VAEs) at UAI 2023.
• Work on stable training of differentiable particle smoothers, for vision and robotics tasks like city-scale global vehicle localization, appears at NeurIPS 2024. This extends our earlier NeurIPS 2023 paper on discriminative particle filters.
• At NeurIPS 2023, work on incorporating graphical models in deep generative models to learn discrete representations.
• A large NSF grant funds research making collaboration accessible for visually impaired workers.
• A new model for sparse graphs with overlapping communities appears at NeurIPS 2022. Related work was presented at the 13th International Conference on Bayesian Nonparametrics.
• The ICML 2021 Time Series Workshop best poster award goes to our work on prediction constraints for semi-supervised classification with Hidden Markov models.
• Our cascaded 3D detection framework, which integrates geometric and contextual cues for robust scene understanding from RGB-D images, is summarized by a 2020 paper appearing in IEEE PAMI.
• An NSF Robust Intelligence Award with Alex Ihler supports work on new particle-based algorithms for inference and learning with continuous graphical models. I gave a talk at the 2017 SoCal Machine Learning Symposium about our earlier diverse particle max-product algorithm, which gives state-of-the-art predictions of continuous protein side-chain conformations.
• At AISTATS 2018, our framework for prediction-constrained training of probabilistic models leads to improved semi-supervised learning of topic models, with applications to the analysis of documents and electronic health records. This work received the SoCal NLP Symposium best paper award.
• An NSF CAREER Award supports our open source toolbox BNPy: Bayesian Nonparametric clustering for Python. BNPy implements scalable, stochastic and memoized variational inference algorithms for Bayesian nonparametric models.
• Work with BrainGate on multiscale semi-Markov dynamics for improved brain-computer interfaces appeared at NeurIPS 2017. A supplemental video demonstrates accurate, interactive control of a computer cursor by a clinical trial participant.
• The 2014 ISBA Mitchell Prize for Bayesian analysis of an important applied problem goes to our NET-VISA system for global seismic monitoring, learned from data provided by the comprehensive nuclear-test-ban treaty organization (CTBTO). For details see the Brown University news article.
• Weiss & Pearl introduce our review article on Nonparametric Belief Propagation for the CACM.

Editorial Highlights

• Scientific committee and local organizing committee for the 14th International Conference on Bayesian Nonparametrics.
• Senior area chair for AISTATS 2025 and area chair for NeurIPS 2025.
• Sponsor chair for the 2018 & 2019 International Conference on Machine Learning.
• Editor, IEEE PAMI Special Issue on Bayesian Nonparametrics, Feb. 2015. (editorial)
• Organizer, ICERM Workshop & Tutorials on Bayesian Nonparametrics, Sept. 2012. (group photo)
• Editor, IEEE Signal Processing Magazine special issue on Recent Advances & Emerging Developments of Graphical Models, Nov. 2010. (editorial)
• Editor, IEEE PAMI Special Issue on Probabilistic Graphical Models in Computer Vision, Oct. 2009. (editorial)

Erik B. Sudderth
E: lastname@uci.edu
P: (949) 824-8169

Office: Donald Bren Hall 4206
Mailing Address:
University of California, Irvine
School of Information & Computer Sciences
Irvine, CA 92697-3435