Hyung-Seok Oh

Education & Experience

Integrated M.S. & Ph.D., Artificial Intelligence

Korea University, Seoul, South Korea

2021 - Present

B.S., Computer Science Engineering

Konkuk University, Seoul, South Korea

2017 - 2021

Machine Learning Researcher Intern

NAVER Cloud, Seongnam, South Korea

Mar 2023 – Jun 2023

News

2026.04
One paper accepted to ACL findings 2026.
2026.01
One paper accepted to ICLR 2026.
2025.10
Received AI Frontier Challenge Excellence Award.
2025.08
One papers accepted to EMNLP 2025.
2025.05
Three papers accepted to INTERSPEECH 2025.
2025.04
One paper accepted to IEEE Transactions on Affective Computing.
2025.04
One paper accepted to IEEE/ACM Transactions on Audio, Speech, and Language Processing.
2025.01
One paper accepted to IEEE Transactions on Affective Computing.
2024.12
One paper accepted to ICASSP 2025.
2024.05
One paper accepted to INTERSPEECH 2024.
2024.04
One paper accepted to IEEE/ACM Transactions on Audio, Speech, and Language Processing.
2023.05
One paper accepted to INTERSPEECH 2023.
2023.03
Worked as Machine Learning Researcher Intern at NAVER Cloud.

Publications

2026

Affectron: Emotional Speech Synthesis with Affective and Contextually Aligned Nonverbal Vocalizations

D.-H. Cho, H.-S. Oh, S.-B. Kim, S.-W. Lee

ACL findings
- Zero-Shot Text-to-Speech
- Emotional Speech Synthesis
- Nonverbal Vocalization Synthesis
Toward Complex-Valued Neural Networks for Waveform Generation

H.-S. Oh, D.-H. Cho, S.-B. Kim, S.-W. Lee

ICLR
- Waveform Generation
- Complex-Valued Neural Networks

2025

FillerSpeech: Towards Human-Like Text-to-Speech Synthesis with Filler Injection and Filler Style Control

S.-B. Kim, J.-H. Cha, H.-S. Oh, H.-J. Choi, S.-W. Lee

EMNLP
- Text-to-speech
- Filler
- Style Control
EmoSphere-SER: Enhancing Speech Emotion Recognition through Spherical Representation with Auxiliary Classification

D.-H. Cho, H.-S. Oh, S.-B. Kim, S.-W. Lee

INTERSPEECH
- Speech emotion recognition
- Dimensional emotion
- Affective computing
VibE-Singer: Vibrato Extraction with High Frequency Pitch for Singing Voice Conversion

J.-S. Choi, D.-M. Byun, H.-S. Oh, S.-W. Lee

INTERSPEECH
- Singing voice conversion
- Singing style transfer
- Vibrato control
- Discrete wavelet transform
DiEmo-TTS: Disentangled Emotion Representations via Self-Supervised Distillation for Cross-Speaker Emotion Transfer in Text-to-Speech

D.-H. Cho, H.-S. Oh, S.-B. Kim, S.-W. Lee

INTERSPEECH
- Text-to-speech,
- Cross-speaker emotion transfer
EmoSphere++: Emotion-Controllable Zero-Shot Text-to-Speech via Emotion-Adaptive Spherical Vector

D.-H. Cho, H.-S. Oh, S.-B. Kim, S.-W. Lee

IEEE Transactions on Affective Computing
- Emotional speech synthesis
- Emotion transfer
- Emotion style and intensity control
- Zero-shot text-to-speech
UnitCorrect: Unit-based Mispronunciation Correcting System with a DTW-based Detection

H.-W. Bae, H.-S. Oh, S.-B. Kim, S.-W. Lee

IEEE/ACM Transactions on Audio, Speech, and Language Processing
- Conditional flow matching, dynamic time
- Dynamic warping
- Mispronunciation detection
- Pronunciation correction
- Speech editing
DurFlex-EVC: Duration-Flexible Emotional Voice Conversion Leveraging Discrete Representations Without Text Alignment

H.-S. Oh, S.-H. Lee, D.-H. Cho, S.-W. Lee

IEEE Transactions on Affective Computing
- Duration control
- Emotional voice conversion
- Self-supervised representation
- Style disentanglement
JELLY: Joint Emotion Recognition and Context Reasoning with LLMs for Conversational Speech Synthesis

J.-H. Cha, S.-B. Kim, H.-S. Oh, S.-W. Lee

ICASSP
- Conversational speech synthesis
- Dialogue, emotional context reasoning
- Large language model
- Low-rank adaptation

2024

EmoSphere-TTS: Emotional Style and Intensity Modeling via Spherical Emotion Vector for Controllable Emotional Text-to-Speech

D.-H. Cho, H.-S. Oh, S.-B. Kim, S.-H. Lee, S.-W. Lee

INTERSPEECH
- Text-to-speech
- Expressive emotional speech synthesis
- Emotional style and intensity control
Diffprosody: Diffusion-based latent prosody generation for expressive speech synthesis with prosody conditional adversarial training

H.-S. Oh, S.-H. Lee, S.-W. Lee

IEEE/ACM Transactions on Audio, Speech, and Language Processing
- Text-to-speech
- Speech synthesis
- Denoising diffusion model
- Prosody modeling
- Generative adversarial networks

2023

HierVST: Hierarchical Adaptive Zero-shot Voice Style Transfer

S.-H. Lee, H.-Y. Choi, H.-S. Oh, S.-W. Lee

INTERSPEECH
- Voice conversion
- Voice style transfer
- Zero-shot voice conversion
- Self-supervised speech representation

2021

Haze removal network using unified function for image dehazing

H.-S. Oh, H.-N. Kim, Y.-J. Kim, C.-H. Yim

Electronics Letters
- Dehaze
- U-Net

Projects

2025 - 2026
Full Duplex Speech-to-Speech Conversational AI

Murf.AI Corp.
2024 - 2025
Research on Interjection Utterances for Natural Speech Synthesis

Samsung Research
2022 - 2023
Development of High-Quality Few-Shot Voice Conversion Technology

NAVER
2021 - 2022
Development of AI Voice Technology for Emotion Controllable Expressive Speech Synthesis

Netmarble

Awards

2025

AI Frontier Challenge - Excellence Award

제1회 (사)한국인공지능학회 AI Frontier Challenge - 우수상

Patents

2025

Emotional Expression Voice Generation Apparatus and Method Capable of Controlling Emotional Style and Intensity Using Continuous Emotional Dimensions

Korean Patent · 10-2908159
2025

Method and Apparatus for Performing Expressive Speech Synthesis via Diffusion-Based Latent Prosody Generation and Prosodic Conditional Adversarial Training

Korean Patent · 10-2809192
2025

Cross-Speaker Emotional Speech Synthesizing Device Using Self-Supervised Emotional Separation Learning and Method of the Same

Korean Patent Application · 10-2025-0130127
2025

Method and Device for Vibrato Controllable Singing Voice Conversion

Korean Patent Application · 10-2025-0111807
2025

Apparatus and Method for Speech Synthesis

Korean Patent Application · 10-2025-0088028
2025

Method for Synthesizing Speech, Recording Medium and Device for Performing Same

Korean Patent Application · 10-2025-0012914
2023

Haze Removal Apparatus and Method Using Unified Function Based on Neural Networks

Korean Patent · 10-2021-0112750

Services

Reviewer

IEEE/ACM Transactions on Audio, Speech, and Language Processing (TASLP), IEEE Transactions on Affective Computing (TAFFC), ACL