Efthymios (Themis) Tsaprazlis
PhD Student | SAIL | University of Southern California
PhD Student | SAIL | University of Southern California
I'm a Computer Science PhD student at University of Southern California and a graaduate research assistant at Signal Analysis and Interpretation Lab (SAIL) , under the supervision of Professor Shrikanth Narayanan. Previously, I did my Bachelors in the ECE Departement at National Technical University of Athens where I worked with Professor Petros Maragos. Also, I was an external collaborator of the Machine Learning Lab at Univerity of Texas at Austin, where I worked with Professor Alex Dimakis .
My research focuses on developing theoretical and practical frameworks for privacy in machine learning, with an emphasis on multimodal and language-based systems. I aim to establish formal definitions of privacy that capture context-sensitive risks and design anonymization methods that balance identity protection with utility and fluency. By creating benchmarks, datasets, and scalable toolkits, I seek to advance principled approaches that both deepen our scientific understanding of privacy and enable deployable, real-world solutions.
• [Jan 2025] We got a paper accepted in ICRA 2025.
• [Aug 2024] I started my PhD at University of Southern California.
• [Jan 2024] I started a new position as Research Assistant in Robotics Unit of the Athena Research and Innovation Center (Athena RC) and CVSP Group.
• [Oct 2023] Our paper "Enhancing CLIP with a Third Modality" got accepted in NeurIPS 2023 Workshop: Self-Supervised Learning - Theory and Practice.
• [Oct 2023] I joined Workable as a Machine Learning Engineer intern.
Towards Open-ended Robotic Exploration using Vision-Inspired Similarity and Foundation Models
Panagiotis P. Filntisis, Efthymios Tsaprazlis, Paraskevas Oikonomou, Francesco Mattioli, Vieri Giuliano Santucci, George Retsinas, Petros Maragos
International Conference on Robotics and Automation (ICRA), 2025
Enhancing CLIP with a Third Modality
Efthymios Tsaprazlis, Georgios Smyrnis, Alexandros G. Dimakis, Petros Maragos
NeurIPS Workshop: Self-Supervised Learning - Theory and Practice, 2023
You can reach me via email at tsaprazl@usc.edu or connect with me on LinkedIn.