Munhak Lee

Researcher, AI center, Samsung Research

Joined

May 2021

Names

Munhak Lee (Preferred)

Mun-hak Lee

Mun-Hak Lee

Emails

****@hanyang.ac.kr (Confirmed)

****@gmail.com (Confirmed)

****@samsung.com (Confirmed)

Personal Links

Career & Education History

Researcher

AI center, Samsung Research (samsung.com)

2025 – Present

PhD student

Electronic Engineering, Hanyang University (hanyang.ac.kr)

2018 – 2025

Advisors, Relations & Conflicts

No relations added

Expertise

speech recognition

2017 – 2025

natural language processing

2017 – 2025

Publications

BBPE16: UTF-16-BASED BYTE-LEVEL BYTE-PAIR ENCODING FOR IMPROVED MULTILINGUAL SPEECH RECOGNITION
Hyunsik Kim, Haeri Kim, Munhak Lee, Kyungmin Lee
- ICASSP
- Readers: Everyone
Bayesian Language Model Adaptation for Personalized Speech Recognition
Mun-Hak Lee, Ji-Hwan Mo, Ji-Hun Kang, Jin-Young Son, Joon-Hyuk Chang
- IEEE Signal Process. Lett. 2025
- Readers: Everyone
Improving Generalization of End-to-End ASR through Diversity and Independence Regularization
Ye-Eun Ko, Mun-Hak Lee, Dong-Hyun Kim, Joon-Hyuk Chang
- INTERSPEECH 2025
- Readers: Everyone
Proper Error Estimation and Calibration for Attention-Based Encoder-Decoder Models
Mun-Hak Lee, Joon-Hyuk Chang
- IEEE ACM Trans. Audio Speech Lang. Process. 2024
- Readers: Everyone
Balanced-Wav2Vec: Enhancing Stability and Robustness of Representation Learning Through Sample Reweighting Techniques
Mun-Hak Lee, Jae-Hong Lee, Do-Hee Kim, Ye-Eun Ko, Joon-Hyuk Chang
- INTERSPEECH 2024
- Readers: Everyone
Whisper Multilingual Downstream Task Tuning Using Task Vectors
Ji-Hun Kang, Jae-Hong Lee, Mun-Hak Lee, Joon-Hyuk Chang
- INTERSPEECH 2024
- Readers: Everyone
Minimum Edit Distance Training for Conditional Language Generation Models
Munhak Lee, Joon-Hyuk Chang
- ICLR 2024 Conference Withdrawn Submission
- Readers: Everyone
Knowledge Distillation From Offline to Streaming Transducer: Towards Accurate and Fast Streaming Model by Matching Alignments
Ji-Hwan Mo, Jae-Jin Jeon, Mun-Hak Lee, Joon-Hyuk Chang
- ASRU 2023
- Readers: Everyone
Cross-Modal Learning for CTC-Based ASR: Leveraging CTC-Bertscore and Sequence-Level Training
Mun-Hak Lee, Sang-Eon Lee, Ji-Eun Choi, Joon-Hyuk Chang
- ASRU 2023
- Readers: Everyone
Regularizing Transformer-based Acoustic Models by Penalizing Attention Weights
Mun-Hak Lee, Joon-Hyuk Chang, Sang-Eon Lee, Ju-Seok Seong, Chanhee Park, Haeyoung Kwon
- INTERSPEECH 2022
- Readers: Everyone

View all 12 publications

Co-Authors

Chanhee Park
Do-Hee Kim
Dong-Hyun Kim
Haeri Kim
Haeyoung Kwon
Hyunsik Kim
Jae-Hong Lee
Jae-Jin Jeon
Ji-Eun Choi
Ji-Hun Kang
Ji-Hwan Mo
Jin-Young Son
Joon-Hyuk Chang
Ju-Seok Seong
Kyungmin Lee
Sang-Eon Lee
Ye-Eun Ko