Jul 2021 - Feb 2022, Sogang Univ.
- Developed a Korean summarization model for a reading activity support application using T5.
- Implemented sentence-length-specific Korean summarization model, achieving a patent.
- Yerin Choi, Hyun Eui Hong, Myoung-Wan Koo, Jeongpil Lee. (2021). Republic of Korea. System for summarizing automatically arbitrary sentences (Patent Number: 10-2021-0152256).
Sep 2022 - Oct 2022, Sogang Univ.
- Built an automated pipeline for detecting and correcting errors in Korean ASR corpora
- Reduced manual annotation effort while significantly improving data consistency
- Won 1st place in the Data-Centric AI Hackathon hosted by National Information Society Agency (NIA) of Korea
- Won the Best Paper Award in Korea Computer Congress 2023
Jun 2022 - Oct 2022, SK Telecom
- Developed a hybrid chatbot combining task-oriented dialogue and chit-chat for personalized lecture recommendations
- Successfully deployed to SK Telecom internal system for production use
- Won 1st place in SKT AI Fellowship
Jun - Jul 2023, SmilegateAI
- Developed a style transfer model for the dialogue tailored to the communication style of women in their 20s.
- Multi-turn dialogue data augmentation
- Deployed real-time chat-bot service
Mar 2022 - Dec 2023, Sogang Univ.
- Supported by Korean government, “Development of Robot and Contents Supporting Children’s Reading Activities Based on AI”.
- Developed an expressive TTS system for children’s reading activity support robot.
- Implemented discrete code-based architecture with mixer models for natural prosody generation.
Relevant Publications:
• [Interspeech 2023] DC CoMix TTS: An End-to-End Expressive TTS with Discrete Code Collaborated with Mixer
Mar 2022 - Dec 2023, Sogang Univ.
- Supported by Korean government, “Development of AI Technology that Provides Dialog-Based Multi-Modal Explainability”.
- Developed a diagnosis and explanation system for dysarthric speech to support clinical therapy applications.
Relevant Publications:
• [ICASSP 2024] Inappropriate Pause Detection in Dysarthric Speech Using Large-Scale Speech Recognition
• [SLT 2024] Speech Recognition-based Feature Extraction for Enhanced Automatic Severity Classification in Dysarthric Speech
Apr 2025 - Ongoing, KIST
- Supported by Korean government, “Development of self-evolving AI bias detection-correction-explain platform based on international multidisciplinary governance”
- Developing knowledge-grounded evaluation frameworks for bias assessment and mitigation in text-to-image and vision-language models.
- Designing human preference datasets to validate evaluation methods and ensure alignment with diverse perspectives on fairness.
Published in Interspeech 2023, 2023
Yerin Choi, Myoung-Wan Koo
Published in ICASSP 2024, 2024
Yerin Choi, Jeehyun Lee, Tae-Jin Song, Myoung-Wan Koo
Published in IEEE Spoken Language Technology Workshop (SLT) 2024, 2024
Yerin Choi, Jeehyun Lee, Myoung-Wan Koo
Published in Workshop on LLMs and KRR for Trustworthy AI @ KR 2025, 2025
Yerin Choi, Seungyeon Ji, Jiwon Kim, Hong Kyu Kim, Kyungjoon Oh, Mi Lim Cheon, Caren Han, Sangwook Yi, Kyungreem Han
Published:
This is a description of your talk, which is a markdown file that can be all markdown-ified like any other post. Yay markdown!
Published:
This is a description of your conference proceedings talk, note the different field in type. You can put anything in this field.
Undergraduate course, University 1, Department, 2014
This is a description of a teaching experience. You can use markdown like any other post.
Workshop, University 1, Department, 2015
This is a description of a teaching experience. You can use markdown like any other post.