Kai-Wei Chang is a third-year Ph.D. student at National Taiwan University, supervised by Professor Hung-yi Lee. His research focuses on speech self-supervised learning and speech language models within the prompting paradigm. Particularly, He focuses on reshaping the speech foundation model into a comprehensive universal speech processing framework.

He has contributed to publications in esteemed speech conferences including ICASSP, Interspeech, ASRU, and SLT. Additionally, he has served as a tutorial speaker at ICASSP 2023, ICASSP 2024 and gained valuable experience as a research scientist intern at Meta’s Reality Labs.

Selected Publications

• SpeechPrompt: An Exploration of Prompt Tuning on Generative Spoken Language Model for Speech Processing Tasks
  Kai-Wei Chang, Wei-Cheng Tseng, Shang-Wen Li, Hung-yi Lee
  Interspeech 2022
  

• SpeechPrompt v2: Prompt Tuning for Speech Classification Tasks
  Kai-Wei Chang, Yu-Kai Wang, Hua Shen, Iu-thing Kang, Wei-Cheng Tseng, Shang-Wen Li, Hung-yi Lee
  Preprint

• SpeechGen: Unlocking the Generative Power of Speech Language Models with Prompts
  Kai-Wei Chang*, Haibin Wu*, Yuan-Kuei Wu*, Hung-yi Lee
  Preprint

  — Based on SpeechPrompt, SpeechPrompt v2 and SpeechGen. A journal paper is submitted to IEEE TASLP. —

• Prompting and Adapter Tuning For Self-Supervised Encoder-Decoder Speech Model
  Kai-Wei Chang, Ming-Hsin Chen, Yun-Ping Lin, Jing Neng Hsu, Paul Kuo-Ming Huang, Chien-yu Huang, Shang-Wen Li, Hung-yi Lee
  ASRU 2024

• Toward degradation-robust voice conversion
  Kai-Wei Chang*, Chien-yu Huang*, Hung-yi Lee
  ICASSP 2022

For the full publication list, please refer to my google scholar

Research Experience

Internship

• Meta Reality Labs - Research Scientist Intern June 2023 - Dec. 2023

Industrial Cooperation

• Meta - Data Mining with Seamless Communication Feb. 2022 - Present

• ASUS - Taiwanese Speech Recognition & Translation Nov. 2021 - Present

• Bonio Inc. - Taiwanese Pronounciation Evaluation System Oct. 2021 - June 2022

Conference Tutorials

• ICASSP 2023: Parameter‑Efficient Learning for Speech and Language Processing: Adapters, Prompts, and Reprogramming
  Pin-Yu Chen (IBM Research), Hung-yi Lee (National Taiwan University), Chao-Han Huck Yang (Georgia Institute of Technology), Kai-Wei Chang (National Taiwan University), Cheng-Han Chiang (National Taiwan University)

• ICASSP 2024: Parameter-Efficient and Prompt Learning for Speech and Language Foundation Models
  Chao-Han Huck Yang (Amazon), Pin-Yu Chen (IBM Research), Hung-yi Lee (National Taiwan University),Kai-Wei Chang (National Taiwan University), Cheng-Han Chiang (National Taiwan University)

Teaching Assistant & Leadership

• Machine Learning, 2021 Spring - Head Teaching Assistant Instrutor: Prof. Hung-yi Lee

  Led and coordinated with over 40 teaching assistants and over 1300 students.

• Linear Algebra, 2020 Fall - Head Teaching Assistant Instrutor: Prof. Hung-yi Lee

  Led and coordinated with over 6 teaching assistants and about 130 students.