01

Know
About me

I am currently a Ph.D. student at the Graduate Institute of Computer Science and Information Engineering of National Taiwan University. I am currently in the Speech Processing Lab of NTU with Prof. Lin-shan Lee and Prof. Hung-yi Lee as my instructor. My research focuses on enhancing the robustness of speech technologies, particularly in developing methods that are resilient in adverse noisy environments and code-switching scenarios. These advancements are applicable to a wide range of speech processing tasks.

02

Publication Overview

2024

• Zero Resource Code-switched Speech Benchmark Using Speech Utterance Pairs For Multiple Spoken Languages (ICASSP 2024)

Kuan-Po Huang, Chih-Kai Yang, Yu-Kuan Fu, Ewan Dunbar, Hung-yi Lee

• Investigating Zero-Shot Generalizability on Mandarin-English Code-Switched ASR and Speech-to-text Translation of Recent Foundation Models with Self-Supervision and Weak Supervision (ICASSP 2024 workshop)

Chih-Kai Yang, Kuan-Po Huang _(co-first), Ke-Han Lu, Chun-Yi Kuan, Chi-Yuan Hsiao, Hung-yi Lee

• Noise Robust Distillation of Self-Supervised Speech Models via Correlation Metrics (ICASSP 2024 workshop)

Fabian Ritter-Gutierrez, Kuan-Po Huang, Dianwen Ng, Jeremy H.M. Wong, Hung-yi Lee, Eng Siong Chng, Nancy F. Chen

• Dataset-Distillation Generative Model for Speech Emotion Recognition (Interspeech 2024 Best Student Paper list)

Fabian Ritter-Gutierrez, Kuan-Po Huang, Jeremy H.M Wong, Dianwen Ng, Hung-yi Lee, Nancy F. Chen, Eng Siong Chng

2023

• Ensemble Knowledge Distillation of Self-Supervised Speech Models (ICASSP 2023)

Kuan-Po Huang, Tzu-hsun Feng, Yu-Kuan Fu, Tsu-Yuan Hsu, Po-Chieh Yen, Wei-Cheng Tseng, Kai-Wei Chang, Hung-yi Lee

2022

• Improving Distortion Robustness of Self-supervised Speech Processing Tasks with Domain Adaptation (Interspeech 2022)

Kuan-Po Huang, Yu-Kuan Fu, Yu Zhang, Hung-yi Lee

• Improving generalizability of distilled self-supervised speech processing models under distorted settings (SLT 2022)

Kuan-Po Huang, Yu-Kuan Fu, Tsu-Yuan Hsu, Fabian Ritter Gutierrez, Fan-Lin Wang, Liang-Hsuan Tseng, Yu Zhang, Hung-yi Lee

2021

• One shot learning for speech separation (ICASSP 2021)

Yuan-Kuei Wu, Kuan-Po Huang _(co-first), Yu Tsao, Hung-yi Lee

• Multi-accent Speech Separation with One Shot Learning (ACL 2021 metaNLP workshop)

Kuan-Po Huang, Yuan-Kuei Wu, Hung-yi Lee

03

My
Experience

• July 2019 - April 2019

  Verizon Media (Yahoo !)

  Search Team Backend Developer Intern

  News Summary App

  • We made a news summary application with Yahoo’s news API.
  • We’ve tried different abstractive and extractive methods. Finally, we designed a sentence ranking system to determine which sentence best fits for being the summary among the whole news document.

• February 2020 - July 2020

  National Taiwan University

  Teacher Assistant

  Course TA

  • Machine Learning 2020 (Prof. Hung-Yi Lee)
  • Deep Learning for Human Language Processing 2020 (Prof. Hung-Yi Lee)
  • Digital Signal Processing 2020 Fall (Prof. Lin-Shan Lee)
  • Machine Learning 2021 (Prof. Hung-Yi Lee)
    • Head Teaching Assistant

• April 11,12,13 2020

  AI Summer School, National Taiwan University

  Teacher Assistant

  Teacher Assistant at AI Summer School

  • I gave a lecture about Source Separation.
  • Hands-on teaching for Source Separation task. colab link
  • AI Summer School 2020 Website

• April 2021

  Interspeech 2021 Paper Reviewer

  Paper Reviewer

  Paper Reviewer of the Interspeech 2021 conference

  • Unofficial Paper Reviewer
  • Interspeech 2021

• May 2021

  ACL 2021 MetaNLP Workshop Paper Reviewer

  Paper Reviewer

  Paper Reviewer of the ACL 2021 MetaNLP Workshop

  • Official Paper Reviewer
  • ACL 2021 MetaNLP Workshop

• August 2021

  ASRU 2021 Workshop Paper Reviewer

  Paper Reviewer

  Paper Reviewer of the ASRU 2021 Workshop

  • Unofficial Paper Reviewer
  • ASRU 2021 Workshop

• September 2022

  ISCSLP 2022 official Paper Reviewer

  Paper Reviewer

  Paper Reviewer of ISCSLP 2022

  • Official Paper Reviewer
  • International Symposium on Chinese Spoken Language Processing 2022

• August 2021

  ICASSP 2023, 2024 Paper Reviewer

  Paper Reviewer

  Paper Reviewer of the ICASSP 2023, 2024 conference

  • Official Paper Reviewer
  • ICASSP 2023
  • ICASSP 2024

04

My
Publications

• Jul 2023 - Sep 2023

  Zero Resource Code-switched Speech Benchmark Using Speech Utterance Pairs For Multiple Spoken Languages

  2024 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP 2024)

  Evaluate the code-switching ability of SSL speech models by language modeling on discrete units. (code)

  Proposed new code-switch speech datasets: (es-en, fr-en, zh-en)

  • Huang, Kuan-Po, et al. "Zero Resource Code-switched Speech Benchmark Using Speech Utterance Pairs For Multiple Spoken Languages." arXiv preprint arXiv:2310.03018 (2023).

• Aug 2022 - Oct 2022

  Ensemble knowledge distillation of self-supervised speech models

  2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP 2023)

  Distil multiple SSL teacher models into one student model.

  • K. . -P. Huang et al., "Ensemble Knowledge Distillation of Self-Supervised Speech Models," ICASSP 2023 - 2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), Rhodes Island, Greece, 2023, pp. 1-5, doi: 10.1109/ICASSP49357.2023.10096445.

• Apr 2022 - Aug 2022

  Improving generalizability of distilled self-supervised speech processing models under distorted settings

  Spoken Language Technology (SLT 2022)

  Self-supervised Speech Processing. Distil HuBERT with noisy speech data.

  • K. -P. Huang et al., "Improving Generalizability of Distilled Self-Supervised Speech Processing Models Under Distorted Settings," 2022 IEEE Spoken Language Technology Workshop (SLT), Doha, Qatar, 2023, pp. 1112-1119, doi: 10.1109/SLT54892.2023.10022474.

• Feb 2022 - Mar 2022

  Improving Distortion Robustness of Self-supervised Speech Processing Tasks with Domain Adaptation

  Conference of the International Speech Communication Association (Interspeech 2022)

  Self-supervised Speech Processing. Continually train HuBERT with noisy data.

  • Huang, Kuan Po, Yu-Kuan Fu, et al. (2022). “Improving Distortion Robustness of Self-supervised Speech Processing Tasks with Domain Adaptation”. In: Interspeech 2022. ISCA-International Speech Communication Association.

• May 2021 - Jul 2021

  Improving the transferability of speech separation by meta-learning

  Speech Separation

  • Huang, Kuan-Po, Yuan-Kuei Wu, and Hung-yi Lee (2021). “Improving the transferability of speech separation by meta-learning”. In: arXiv preprint arXiv:2203.05882.

• Mar 2021 - May 2021

  Multi-accent Speech Separation with One Shot Learning

  Meta Learning and Its Applications to Natural Language Processing Workshop (ACL 2021 workshop)

  Speech Separation

  • Huang, Kuan Po, Yuan-Kuei Wu, and Hung-yi Lee (Aug. 2021). “Multi-accent Speech Separation with One Shot Learning”. In: Proceedings of the 1st Workshop on Meta Learning and Its Applications to Natural Language Processing. Online: Association for Computational Linguistics, pp. 59–66. URL: https://aclanthology.org/2021.metanlp

• Oct 2020 - June 2021

  One Shot Learning for Speech Separation

  2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP 2021)

  Speech Separation

  • Wu, Y. K., Huang, K. P., Tsao, Y., & Lee, H. Y. (2021, June). One shot learning for speech separation. In ICASSP 2021-2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) (pp. 5769-5773). IEEE.

05

My
Projects

06

Contact Me

Email

gerber861017@gmail.com

Social Network