CV
The PDF will be easier to read. For the publications, check PDF or "Publications" menu.
Basics
| Name | Catherine Hae Seung Jeon |
| haeseungjeon@ewha.ac.kr | |
| Phone | (+1) 765-543-3798 |
Interests
| AI for Real-World Applications | |||||
| AI for Security | |||||
| AI for Healthcare | |||||
| AI for Network Analysis | |||||
| Trustworthy AI | |||||
| AI Generalization and Robustness | |||||
| n-shot Learning | |||||
| Continual Learning | |||||
| Metric Learning | |||||
| Data Augmentation | |||||
Education
Research experience
-
2026.01 - Present Visiting Researcher
Purdue University | Sensor, Electro-Acoustics Technology Lab
- Building a physical basis prediction model based on electrical-mechanical coupling, utilizing impedance-based sensing data from electroacoustic edge devices.
-
2025.06 - 2025.08 Visiting Researcher
UTEP | Intelligent and Quantum Secure Advanced Cyber Defense Lab
- Built a continual learning framework with a conditional tabular GAN and improved storage/training efficiency by generative replay. Drove experiments and achieved up to 16% improvements over the baselines (GAN and TVAE) in Class-IL/Time-IL scenarios.
-
2025.05 - 2026.01 Part-Time Researcher
Purdue University | Sensor, Electro-Acoustics Technology Lab
- Directed Transformer-based speech clarity prediction model development to support hearing aid users. Suggested fine-tuning an ASR backbone with speech-in-noise audio and masked transcription for inaudibility simulation, and achieved 5th place in the challenge.
-
2024.09 - Present Graduate Research Assistant
Ewha Womans University | AI Security Lab
- Proposed a robust few-shot model using LLM-inspired normalization (pre-norm, RMSNorm) and dynamic window-based augmentation, achieving up to 9% gains over SimCLR-based SOTA across few-shot, concept drift, and open-set settings without large-scale pre-training.
- Leading advanced time-series embedding model development utilizing metric learning models and FENs in collaboration with the University of Edinburgh and the U.S. NRL.
-
2023.01 - 2024.08 Undergraduate Research Assistant
Ewha Womans University | AI Security Lab
- Led research on the first DL-based network traffic fingerprinting model, achieving a 41-55% gains over a previous SOTA. Built a Selenium crawler to collect 300K+ fresh samples over 1+ years, profiled unique data patterns, and benchmarked ML/DL baselines on a new dataset.
Industry experience
-
2023.03 - 2023.06 Cloud Engineer Intern Seoul, South Korea
Samyang Data Systems Inc. | Cloud Solutions Team
- Implemented a Golang monitoring agent and E2E pipeline that collects, stores, and visualizes real-time metrics from AWS infrastructure by utilizing InfluxDB and Grafana. Deployed agents to 10+ client servers, ensuring software reliability through stress tests.