• Microsoft Research New England: “Privacy, Utility, and Invariant Representations via Adversarial Learning,” invited talk, Machine Learning Discussion Group, 21 October, 2019.
• Amrita Vishwa Vidyapeetham, Amritapuri, India: “Linear Algebra for Machine Learning,” invited lectures, EE and CS Departments, 21-23 December, 2017.
• MIT Lincoln Laboratory: “A Topic Modeling Approach to Learning Preference-Behavior from Pairwise Comparisons,” invited seminar talk (advanced natural language processing theme), 19 May, 2017.
• Amrita Vishwa Vidyapeetham, Amritapuri, India: “Privacy-Preserving Indoor Activity Analysis,” invited talk, EE and CS Departments, 23 December, 2016.
• Amrita Vishwa Vidyapeetham, Amritapuri, India: “A Geometric Approach to Topic Modeling,” invited talk, EE and CS Departments, 22 December, 2016.
• Institut Henri Poincaré, Paris, France: “On the Ultimate Limit of Two-Terminal Interactive Computing,” invited talk in thematic program on Nexus of Information and Computation Theories, Distributed Computation and Communication Theme, Paris, France, 11 Feb., 2016.
• University of Illinois, Urbana–Champaign: “Efficient Distributed Topic Modeling with Provable Guarantees,” invited talk in special session on Information Processing for Big Data in the 53rd Annual Allerton Conference on Communication, Control, and Computing, Monticello, IL, 29 Sep. – 2 Oct., 2015.
• University of Wisconsin, Madison: “A Topic Modeling Approach to Learning Preference-Behavior from Pairwise Comparisons,” invited talk in the Systems, Information, Learning and Optimization (SILO) Seminar Series, ECE Department, 30 September, 2015.
• Stanford University: “A Topic Modeling Approach to Learning Preference-Behavior from Pairwise Comparisons,” invited talk in the Information Theory Forum, EE Department, 22 May, 2015.
• University of California, San Diego: “A Topic modeling approach to ranking from pairwise preferences,” invited talk in the Graduate Seminar in Communication Theory and Systems, ECE Department, 13 May, 2015.
• Harvard University: “A Topic Modeling Approach to Ranking,” invited talk in the EconCS Seminar Series, Computer Science and Economics Departments, 13 Mar., 2015.
• Arizona State University: “A New Geometric Approach to Topic Modeling and Discovery,” invited talk in the Network Science Seminar Series, ECEE Department, 31 Oct., 2014.
• University of California, Berkeley: “A New Geometric Approach to Topic Modeling and Discovery,” invited talk in the Networking, Communications, and DSP Seminar Series, EECS Department, 12 Aug., 2014.
• University of California, San Diego: “A Zero-One Law for Secure Two-Party Sampling,” invited talk in the 9th IEEE International Workshop on Information Theory and Applications (ITA), San Diego, CA, 9– 14 Feb., 2014.
• ParisTech, France: “Interactive Source Coding and Computation : Benefits, Ultimate Limits, and Challenges,” keynote in the GdR ISIS Workshop on Information Theoretic and Coding aspects for Distributed Source Coding and Storage in Networks, Paris, France, 28 Nov., 2013.
• EPFL, Switzerland: “A New Geometric Approach to Topic Modeling and Discovery,” invited talk in the Information Processing Group (IPG) Seminar Series, 27 Nov., 2013.
• University of California, San Diego: “An Impossibility Result for High-Dimensional Supervised Classification,” invited talk in the 8th IEEE International Workshop on Information Theory and its Applications (ITA), San Diego, CA, 11– 15 Feb., 2013.
  
• Harvard University: “Sensing-Aware Inference with High-Dimensional Data,” invited talk in the EE Seminar Series of the Division of Engineering and Applied Sciences, 14 Sep., 2012.
• SPCOM 2012: “Impact of Sensing Structure in Supervised Classification of High-Dimensional Data,” invited talk in the IEEE International Conference on Signal Processing and Communications (SPCOM), Bangalore,India, 22-25 Jul., 2012.
• Banff International Research Station for Mathematical Innovation and Discovery: “The Infinite Message Limit of Interactive Source Coding,” International Workshop on Interactive Information Theory, Banff, Canada, 15–20 Jan., 2012.
• University of Massachusetts, Amherst: “Recognizing Human Actions using Feature Covariance Matrices,” invited talk in the ECE Department Seminar Series, 17 Oct., 2011.
• Banff International Research Station for Mathematical Innovation and Discovery: “Exploring Function and Distribution Structure in Interactive Computing Through Examples,” International Workshop on Algebraic Structure in Network Information Theory, Banff, Canada, 14–19 Aug., 2011.
• Worcester Polytechnic Institute (WPI): “Action Recognition in Covariance Manifolds,” invited talk in the ECE Department Graduate Seminar Series, 17 Feb., 2011.
• Tufts University: “Action Recognition on Covariance Manifolds,” invited talk in the ECE Department Colloquium, 25 Jan., 2011.
• University of California, San Diego: “Kaspi’s Two-Way Source Coding Question Resolved,” invited talk in the 5th IEEE International Workshop on Information Theory and its Applications (ITA), San Diego, CA, 31 Jan.– 5 Feb., 2010.
• Boston University: “Condensing Video and Recognizing Actions,” invited talk in the Department of Computer Science, Image and Video Computing Weekly Symposium, 29 Oct., 2009.
• Boston University: “An Information Processing Tour of Sensor Networks: Field-Estimation, Interactive-Computing, and Visual-Analytics,” tenure talk in the ECE Department Seminar Series, 09 Sep., 2009.
• University of Wisconsin, Madison: “Interactive Function Computation: A Distributed Block Source Coding Perspective,” invited talk in the ECE Department Systems Seminar Series, 12 Mar., 2009.
• University of Illinois, Urbana–Champaign: “Interactive Function Computation: A Distributed Block Source Coding Perspective,” invited talk in the Topics in Systems Seminar Series, Coordinated Science Laboratory, 10 Mar., 2009.
• Harvard University: “Distributed Source Coding for Interactive Function Computation,” invited talk in the EE Seminar Series of the Division of Engineering and Applied Sciences, 13 Feb., 2009.
• University of California, San Diego: “Bounds for Interactive Computation in Collocated Networks,” invited talk in the 4tℎ IEEE International Workshop on Information Theory and its Applications (ITA), San Diego, CA, 8–13 Feb., 2009.
• MIT: “Distributed Source Coding for Interactive Function Computation,” invited talk in the RLE Signal Transformation and Information Representation Seminar Series, 1 Dec., 2008.
• Yale NSF Workshop 2008: “Distributed Coding for Interactive Computing,” invited talk in the US National Science Foundation Workshop on the Frontiers in Distributed Communication, Sensing and Control, Yale University, New Haven, CT, 31 Oct. – 2 Nov., 2008.
• Boston University: “Information-Theoretic Performance Limits and Algorithms for Signal Processing Applications,” invited talk in the BU CISE Faculty Research Workshop, Boston, MA, 11 Jan. 2008.
• University of California, San Diego: “Distributed Field Reconstruction in Sensor Networks,” Center for Advanced Network Sciences (ANS) inaugural invited summer lecture series (2-lecture short-course), California Institute of Telecommunications and Information Technologies (CALIT2), San Diego, CA, 1–10 Aug., 2007.
• Harvard University: “The Benefit of Delays in Sequential Coding of Correlated Sources,” invited talk in the EE Seminar Series of the Division of Engineering and Applied Sciences, 23 Feb. 2007.
• Boston University: “Enabling Low-Power Video-Over-Wireless: Do We Need a Paradigm Shift Based on Distributed Coding?” invited talk in the ECE Department Research Spotlight Seminar Series, 21 Oct. 2005.
• University of Wisconsin, Madison: “Information processing in distributed sensor systems: Data acquisition and Inference,” invited talk in the ECE Department Systems Seminar Series, 6 Apr. 2004.
• Boston University: “Information processing in distributed sensor systems: Data acquisition and Inference,” invited talk in the ECE Colloquium Series, 24 Mar. 2004.
• University of Minnesota, Minneapolis: “Information processing in distributed sensor systems: Data acquisition and Inference,” invited talk in the Digital Technology Center Seminar Series, 12 Mar. 2004.
• University of California, Berkeley: “Distributed Sampling of Smooth Fields,” invited talk in the EECS Graduate Seminar Series on Conceptual Issues in Large-scale Ad-hoc Networks, 3 Nov. 2003.
• University of California, Berkeley: “A Unified Framework for Image Modeling and Estimation using Measurement Constraints,” invited talk in the EECS Networking, Communications, and DSP Systems Seminar Series (sponsored by Cisco), 8 Oct. 2002.
• University of Michigan, Ann Arbor: “Statistical Signal Modeling and Estimation using Multiple Wavelet Bases and the Maximum Entropy Principle,” invited talk in the EECS Communications and Signal Processing Laboratory Seminar Series, 13 Mar. 2002.
• University of California, Santa Barbara: “Statistical Signal Modeling and Estimation using Multiple Wavelet Bases and the Maximum Entropy Principle,” invited talk in the ECE Communications and DSP Seminar Series, 8 Mar. 2002.