Sunjun Hwang
Undergraduate Researcher, School of Computer Science, Yonsei University
Quantum Computing & AI Security Research
About Me
I am an undergraduate researcher majoring in Computer Science at Yonsei University, focusing on quantum computing as my primary research area. My academic interests span quantum algorithms, AI security, and AI hardware accelerators. I am particularly interested in applying quantum-mechanical principles to computational systems and developing secure, efficient AI models.
Research Interests
Quantum Computing
Quantum algorithms, quantum circuit optimization
AI Security
Adversarial attacks & defenses, model robustness
AI Semiconductors
AI accelerators, neuromorphic computing
Quantum–ML Integration
Quantum machine learning, hybrid quantum-classical systems
Carla Autonomous Driving
High-fidelity driving simulation, sensor configuration, scenario generation
TransFuser: Multimodal fusion (RGB+LiDAR), end-to-end waypoint prediction
InterFuser: Hierarchical multimodal transformer, interpretable driving control
Education
2022 - Present
Bachelor of Science
Yonsei University, School of Computer Science
Relevant coursework: Calculus(I,II), General Physics(I,II), General Chemistry(I,II), Computational Thinking, Computer Programming, Data Science, Java Programming,
Data Structures, Algorithms, Opensource and Linux Systems, Object Oriented Programming, Discrete Mathematics, Linear Algebra, System Programming,
Database, Signal and Systems, Artificial Intelligence, Natural Language Processing, Quantum Mechanics(I,II), AI Mathematics, Cryptography,
Microprocessor, Operating System, Software Engineering, Datamining, Graph Theory, AI Security
Publications
Adversarial Robustness Analysis of Deep Learning-Based Automatic Modulation Classification in Wireless Communication
IEEE ICAIIC 2026 (Accepted, Dec 21, 2025)
This paper investigates the adversarial robustness of deep learning–based automatic modulation classification systems in wireless communication environments. Various attack scenarios and robustness evaluation metrics are analyzed to assess model reliability under adversarial perturbations.
Design and Implementation of an FPGA-Based Real-Time Voice Risk Detection System
KCS 2026 (Accepted, Dec 15, 2025)
This work presents the design and implementation of a real-time voice risk detection system on FPGA hardware, focusing on low-latency signal processing and robust emergency voice and scream detection in practical environments.
Quantum-Secured Hybrid Communication System for Tactical Military Networks: Implementation and Performance Analysis of BB84 Protocol Based on PennyLane
Journal of the Korean Institute of Communications and Information Sciences (JKICS), 2026 (Accepted, Dec 01, 2025)
This paper proposes a quantum-secured hybrid communication system for tactical military networks. The BB84 quantum key distribution protocol is implemented using PennyLane, and its performance is analyzed in realistic communication scenarios.
Quantum Noise-based Adversarial Attack on Diffusion Models and Analysis of Defense Mechanisms
KIIT-JICS 2026 (Accepted, Nov 08, 2025)
This study explores a novel adversarial attack framework on diffusion models using quantum noise characteristics and evaluates defense mechanisms against such quantum-inspired perturbations.
Logit-based Knowledge Distillation for Heterogeneous Medical Image Federated Learning
Proceedings of KIIT Conference, 2025 (Accepted, Nov 03, 2025)
This paper proposes a logit-based knowledge distillation approach to improve performance in heterogeneous federated learning environments for medical image analysis while preserving data privacy.
Post-hoc Defense with Knowledge Distillation in Federated Learning: An Empirical Study against FGSM and PGD Attacks
Proceedings of KICS Conference, 2025 (Accepted, Oct 21, 2025)
This work presents a post-hoc defense strategy using knowledge distillation to enhance adversarial robustness in federated learning systems against FGSM and PGD attacks.
Classification of Pneumonia in Chest X-rays Using a Hybrid Neural Network Based on a 3-Qubit Quantum Circuit
KSII Conference, 2025 (Accepted, Sep 23, 2025)
This paper introduces a hybrid quantum–classical neural network incorporating a 3-qubit quantum circuit for pneumonia classification from chest X-ray images.
Performance Comparison of 8 Deep Learning Models for Seismic Signal Denoising
Proceedings of KIIT Conference, 2025 (Sep 05, 2025)
The paper conducts a systematic comparison of eight deep learning models—BiLSTM, Denoising Autoencoder, FFTformer, Informer, PatchTST, TCN, UNet1D, and WaveNet—for seismic signal denoising under identical experimental conditions.
A Study on Robustness Enhancement and Multi-Adversarial Attacks in Vision Transformer-based Image Classification Models
Proceedings of KIIT Conference, 2025 (May 16, 2025)
This paper analyzes adversarial robustness in ViT-B32 models under FGSM, PGD, and CW attacks, demonstrating that multi-adversarial training significantly improves robustness while maintaining high clean accuracy.
Projects
OCR Project
April 2025 - November 2025
Project description: This project strengthened my ability to apply classical computer vision and OCR techniques to real-world data, and taught me the importance of designing systems that are resilient to noise and variability in user-generated documents.
Naver OCR
OpenAI API
Ubuntu
Development of a Hazard Detection System Using Emergency Voice and Scream Detection
Sep 2024 - Dec 2024
Project description: This project demonstrated the power of combining time-frequency analysis with modern neural architectures for real-world safety applications. It strengthened my practical understanding of signal processing, deep learning optimization, and team coordination in technical research environments.
PyTorch
Torchvision
Scipy
Librosa
Google Speech-to-Text API
User-Recommended Search Term Program
April 2024 - June 2024
Project description: At the time, I had no exposure to machine learning or deep learning, but I learned how far simple algorithms and thoughtful design can go. This experience taught me to build solutions with the tools I had, and to always design with the user in mind.
PyTorch
Python
Side Channel Attack
March 2025 - June 2025
Project description: This project demonstrated the practical viability of deep learning in detecting side-channel anomalies, even under simulated conditions. Leading a research-focused, multi-member effort strengthened my ability to manage complex technical collaborations and clearly communicate experimental outcomes. The MLP model’s efficiency and robustness underlined the importance of model–data alignment over architectural complexity. Moving forward, we plan to integrate real-world trace data, apply advanced preprocessing techniques, and experiment with adversarial robustness to build deployable side-channel defense systems.
PyTorch
Python
Scikit-learn
PCA
Technical Skills
Python
C++
Qiskit
Cirq
PyTorch
TensorFlow
CUDA
Quantum Computing
Machine Learning
Deep Learning
Awards & Honors
2023-2
Academic Excellence Award
Yonsei University
2024-1
Academic Excellence Award
Yonsei University
2024
Software Engineer Literacy Seminar
The Korean Institute of Convergence Signal Processing
ongoing
Volunteering and social contributions
The Korean Red Cross
2024
LG Aimers Certificate of Achievement
LG
2024
Certificate of Completion of the Artificial Intelligence Convergence Technology Expert Training Program
Korean Artificial-Intelligence Convergence Technology Society
2025 07
Outstanding Paper Award from the Korea Information Technology Society
KIIT
2025 12
Outstanding Paper Award from the Korea Information Technology Society
KIIT