cv

Yi-Cheng Lin — PhD student in Communication Engineering at National Taiwan University. Research on fairness and social bias in speech technology, speech emotion recognition, and evaluation of large audio-language models.

Contact Information

Name Yi-Cheng Lin
Professional Title PhD Student in Communication Engineering, National Taiwan University
Email f12942075@ntu.edu.tw
Phone +886 975-745-477
Location Taipei, Taiwan
Website https://dlion168.github.io

Professional Summary

Ph.D. student at National Taiwan University working on large audio-language models and fairness in speech processing, advised by Hung-yi Lee in the Speech Processing and Machine Learning Lab. 30+ papers at Interspeech, ICASSP, ASRU, SLT, ACL, EMNLP and IEEE TASLP with over 700 citations. Best Paper Runner-Up at Interspeech 2024.

Experience

  • 2026 - Present
    Student Researcher
    Google Cloud AI
  • 2022 - 2023
    Research & Development Intern
    Microsoft Taiwan
    • Upgraded the geolocation prediction system from rule-based to ML-based, improving accuracy by 2% and reducing prediction time by 10%.
  • 2022 - 2022
    Intern
    ASUS AICS

Education

  • 2025 - Present

    Taipei, Taiwan

    PhD
    National Taiwan University
    Communication Engineering
    • Speech Processing and Machine Learning Lab, advised by Hung-yi Lee
    • Selected courses: Reinforcement Learning, Deep Learning for Human Language Processing
  • 2023 - 2025

    Taipei, Taiwan

    M.S.
    National Taiwan University
    Communication Engineering
    • Transferred to the PhD program in Spring 2025
  • 2018 - 2023

    Taipei, Taiwan

    B.S.
    National Taiwan University
    Electrical Engineering
    • Double major in the Trans-disciplinary Bachelor Program (2021.09–2023.06)
    • Minor in Physics
    • Completed the Social Design, Leadership, and Quantum Computation & Quantum Information programs

Awards

  • 2026
    First Place, Track 3
    VoiceMOS Challenge 2026

    Speaker and accent similarity prediction for codec-based speech synthesis.

  • 2026
    Foundation for the Advancement of Outstanding Scholarship Travel Grant
    Foundation for the Advancement of Outstanding Scholarship

    For attending ACL 2026.

  • 2026
    ICASSP 2026 Travel Grant
    IEEE Signal Processing Society
  • 2025
    Foundation for the Advancement of Outstanding Scholarship Travel Grant
    Foundation for the Advancement of Outstanding Scholarship
  • 2025
    INTERSPEECH 2025 Travel Grant
    ISCA
  • 2025
    First Place, Track 1 and Track 3
    ASRU AudioMOS Challenge 2025

    MOS prediction for text-to-music systems, and for speech at high sampling rates.

  • 2025
    NSTC Graduate Research Fellowship
    National Science and Technology Council, Taiwan
  • 2024
    Garmin Scholarship Award
    Garmin
  • 2024
    Best Paper Runner-Up in the Responsible Speech Foundation Models Special Session
    INTERSPEECH 2024

    For “On the social bias of speech self-supervised models.”

  • 2023
    Irving T. Ho Memorial Scholarship
    Irving T. Ho Memorial Foundation
  • 2023
    Outstanding Engineering Student Scholarship
    Chinese Institute of Engineers
  • 2021
    Pan Wen-Yuan Foundation Student Scholarship
    Pan Wen-Yuan Foundation

Publications

Service

Reviewer: Conferences — ACL Rolling Review (Feb 2024, Oct 2024, Feb 2025, May 2025), INTERSPEECH 2025, ASRU 2025, ACML 2025, ICLR 2025, ROCLING 2025, ICASSP 2026. Journals — IEEE Transactions on Audio, Speech and Language Processing; IEEE Transactions on Affective Computing.
Teaching Assistant: National Taiwan University — Deep Learning for Human Language Processing (Fall 2023, Spring 2026); Special Project (every semester since Fall 2023).

Skills

Research Areas: Fairness and Social Bias in Speech Technology, Speech Emotion Recognition, Large Audio-Language Model Evaluation, Speech Quality Assessment, Self-Supervised Speech Representation Learning, Neural Audio Codecs, Speech and Music Representation Learning
Technical: PyTorch, Speech and audio processing, Large-scale model training and distillation, Benchmark and dataset construction, Subjective and objective evaluation design

Languages

Mandarin Chinese : Native speaker
Taiwanese Hokkien : Conversational
English : Professional working proficiency

Interests

Fairness in Speech AI: Demographic disparity in speech models, Accent and first-language bias, Debiasing without demographic labels