Research
Topological Guidance-Based Knowledge Distillation (ICML 2024)
I develop topology-aware model compression methods using persistent homology to guide student networks.
Poster and paper presented at ICML 2024.
Download Paper
Code
Show Abstract
A lightweight knowledge-distillation framework that leverages topological features from multiple teachers to enhance model performance and robustness while avoiding the computational cost of traditional TDA.
Los Alamos National Laboratory – Dynamic Summer School (2024)
Robust AI workflows for critical infrastructure applications.
Poster presented at LANL DSS 2024.
Download Poster
Show Abstract
This work investigates ways to enhance AI workflows for critical infrastructure.
Arizona State University Research Symposium (2024)
Time-aware representation learning for gait analysis.
Poster presented at ASU’s annual research symposium.
Download Poster
Show Abstract
This work analyzes gait patterns using time-aware neural architectures
Vision-Based State Estimation of Serial Manipulator (IMAC 2026)
Accepted for presentation at IMAC 2026.
Download Paper
Show Abstract
Vision-Based Robotic Arm State Estimation. Using a transfer learning approach, sim-to-real efficiency is achieved.