About Me

I am an assistant professor at the Department of Computer Science at Vanderbilt University. My research is motivated by the challenges facing today’s machine learning (ML) algorithms for deployment in critical physical world applications. My research lab, Machine Intelligence and Neural Technologies (MINT), focuses on creating the next generation of core machine learning algorithms that are mathematically grounded, uncertainty aware, label-efficient, and can continually learn from the nonstationary and never-ending data stream. Among recent directions that my team and I have contributed to are: 1) single and multi-agent lifelong learning machines as part of three Defense Advanced Research Projects Agency (DARPA) programs, 2) uncertainty-aware deep learning as part of the DARPA Enabling Confidence program, 3) geometric deep learning, and 4) applications of optimal mass transportation theory in ML and computer vision. I am an inventor of over 20 issued patents and have over 50 published articles in high-impact machine learning, computer vision, and signal processing journals and conferences.

[Open Positions] I am recruiting motivated PhD students with strong mathematical background to work with me on topics aligned with my lab's 'research directions'. Master’s and undergraduate students within Vanderbilt University and visiting scholars are always welcome. Please refer to my lab's website for more information.

News

• Congratulations to Huy Tran and the team for acceptance of their paper, "Stereographic Spherical Sliced Wasserstein Distances," to ICML 2024.
• Our collaborative paper on "A Collective AI via Lifelong Learning and Sharing at the Edge" is now published at Nature Machine Intelligence.
• Congratulations to Ali Abbasi for acceptance of his paper, "BrainWash - A Poisoning Attack to Forget in Continual Learning" to CVPR 2024
• Congratulations to Dr. Yikun Bai for acceptance of his paper at CVPR 2023.
• Congratulations to my students Zihao (Harry) Wu (CS Undergrad) and Huy Tran for acceptance of their paper to IEEE ICASSP 2023.
• Congratulations to my student Ali Abbasi for acceptance of his paper to the Proceedings of the Conference on Lifelong Learning Agents, 2022.
• Our paper on "Generalized Sliced Probability Metrics'' received the Best Paper Award from IEEE ICASSP 2022 - May 27, 2022
• Our Machine Learning Seminars are now available on YouTube! - Feb 20, 2022
• I gave a talk at the One World Seminar on the Mathematics of Machine Learning focused on "Wasserstein Embeddings." (Video) - Nov 24, 2021.
• Our paper 'Pooling by Sliced Wasserstein Embedding' got accepted to NeurIPS 2021 - Oct 06, 2021.
• I joined Vanderbilt University as an Assistant Professor of Computer Science - Aug 16, 2021.
• Our paper 'Wasserstein Embedding for Graph Learning' got accepted to ICLR2021 for a poster presentation - Jan 07, 2021.
• Our paper 'Statistical and Topological Properties of Sliced Probability Divergences' got accepted to NeurIPS2020 for a spotlight presentation - Sep 25, 2020.

• We presented our paper 'SAR Image Classification Using Few-Shot Cross-Domain Transfer Learning' in CVPRW'19 (Oral presentation), Long Beach, CA, USA - June 16 2019.
• Our paper on 'Deep Transfer Learning for Few-Shot SAR Image Classification' got accepted to the IEEE Journal of Remote Sensing.
• We presented our 'Sliced-Wasserstein Auto-Encoder' paper in ICLR'19, New Orleans, LA, USA - May 9, 2019.
• I gave a talk on 'Optimal Transport in Biomedical Imaging' in the British Applied Mathematics Colloquium 2019 (BAMC'19), at Unviersity of Bath, UK - April 25, 2019. (slides)
• I gave a talk on 'Generalized Sliced-Wasserstein Distances' in the Department of Applied Mathematics at University of Cambridge, UK - April 23, 2019.
• I gave an ECE Graduate Seminar talk at Carnegie Mellon University on Feb 14, 2019, on the topic of "Generalized Sliced-Wasserstein Distances and Their Applications in Generative Modeling and Transfer Learning".
• Our paper "Sliced Wasserstein Auto-Encoders" got accepted to ICLR'19 - Dec 21, 2018
• Our paper "Discovering Molecular Functional Groups Using Graph Convolutional Neural Networks" is now available on arXiv - Dec 6, 2018
• Our proposal titled, 'Super-Turing Evolving Lifelong Learning ARchitecture (STELLAR)', was funded by DARPA. Dr. Hava Siegelmann is the program manager leading the Lifelong Learning Machines (L2M) program at DARPA. The HRL team is led by Dr. Praveen Pilly and I and consists of academic members from six world-renowned universities - July 2018
• We are presenting our paper "Multi-Agent Distributed Lifelong Learning for Collective Knowledge Acquisition" at AAMAS2018 - July 2018
• We are presenting two papers at CVPR2018 paper 1 paper 2 - June 2018
• I received my second IR&D Research Award at HRL Laboratories for our Deep Sense Learning (DSL) project - June 2018
• Our tutorial on "Optimal Transport in Biomedical Imaging" at the IEEE International Symposium on Biomedical Imaging (ISBI) was an absolute success.
• We are presenting our paper "Joint Dictionaries for Zero-Shot Learning" at AAAI'18 - February 2018