I'm a junior at Duke Kunshan University studying Applied Mathematics and Computer Science. My research focuses on robust speech and machine learning systems, particularly speech enhancement, automatic speech recognition, and on-device/federated learning. I'm interested in building practical ML systems that remain reliable under real-world constraints.
- MobileRLHF — Federated preference learning for on-device LLM adaptation. Accepted to the Work-in-Progress (WiP) track at ACM MobiCom 2026.
- FedHealthAgent — Federated and on-device learning for personalized health agents. Under review at ACM MobiCom 2027.
- SE–ASR Artifact-Aware Post-processing — Research on reducing the mismatch between pretrained speech enhancement systems and automatic speech recognition through artifact-aware post-processing. Ongoing research.
- Canine Respiratory Pathogen Surveillance — Data processing and analysis for a molecular surveillance study of respiratory pathogens and coinfections in symptomatic dogs in Kunshan, China. Under review at The Journal of Veterinary Medical Science.
- FedCampus 2.0 — Backend and infrastructure development for a federated campus platform.
- Slouch Watch — A macOS posture-tracking app using AirPods motion sensors for real-time posture monitoring.
- Dear — A mobile app delivering daily supportive messages through notifications and widgets.
Speech Enhancement · Automatic Speech Recognition · Robust Speech Processing · ML Systems · Federated Learning · On-device AI