Affiliation
About
PhD candidate in Intelligent Systems Engineering department in Luddy School of Informatics, Computing, and Engineering at Indiana University Bloomington. I investigate efficient machine learning and deep learning solutions for audio processing problems with my advisor Prof. Minje Kim. My current research focuses on novel modes of compression for source separation applications through personalization and scalable learning for practical deployment onto resource-constrained devices. Previously, I worked on various modes of compression including extreme quantization, hashing, and pruning.I was fortunate enough to work as an applied scientist intern at Amazon Lab126's Audio Technology group and as an engineering intern at Qualcomm's Audio Algorithms group.
I graduated with a BS in Physics from the University of Illinois at Urbana-Champaign.
Please check out my CV for more information.
In my leisure time, I mostly spend time with my two dogs and hiking at nearby parks or nature trails.
News
- 07/14/2021: Our paper, Test-Time Adaptation Toward Personalized Speech Enhancement: Zero-Shot Learning with Knowledge Distillation, was accepted for publication in WASPAA 2021. [session] [paper]
- 06/02/2021: A paper I co-authored, Personalized speech enhancement through self-supervised data augmentation and purification, was accepted for publication in Interspeech 2021.
- 10/20/2020: Accepted return offer for Applied Scientist internship at Amazon Lab126 for Summer 2021!
- 05/05/2020: Presented a virtual oral presentation of our Boosted Locality Hashing paper for ICASSP 2020. [session] [paper]
- 01/24/2020: Our paper, Boosted Locality Sensitive Hashing, was accepted for publication in ICASSP 2020 and nominated as the Best Student Paper.
- 01/23/2020: Applied scientist internship offer accepted at Amazon Lab126 for Summer 2020.
- 08/16/2019: I concluded my exciting internship at Qualcomm for Summer 2019. It was an amazing experience collaborating with the Audio Algorithms team working on nonlinear echo cancellation.
- 05/17/2019: Attended ICASSP 2019 in Brighton, UK. Presented on bitwise recurrent neural networks. [poster] [paper]
Publications
- Sunwoo Kim, Minje Kim, "Test-Time Adaptation Toward Personalized Speech Enhancement: Zero-Shot Learning with Knowledge Distillation," in IEEE Workshop on Applications of Signal Processing to Audio and Acoustics (WASPAA), New Paltz, NY, Oct 17-20, 2021. [Demo and Presentation, Code]
- Aswin Sivaraman, Sunwoo Kim, Minje Kim, "Personalized Speech Enhancement through Self‑Supervised Data Augmentation and Purification," in Interspeech, Brno, Czech Republic, Aug 30-Sep 3, 2021.
-
Sunwoo Kim, Haici Yang, Minje Kim, "Boosted Locality Sensitive Hashing: Discriminative Binary Codes For Source Separation," in Proceedings of the International Conference on Acoustics, Speech, and Signal Processing (ICASSP), Barcelona, Spain, May 4-8, 2020. [Demo, Code, Presentation]
[Nominated as the Best Student Paper]
- Sunwoo Kim, Mrinmoy Maity, Minje Kim, "Incremental Binarization On Recurrent Neural Networks For Single-Channel Source Separation," in Proceedings of the International Conference on Acoustics, Speech, and Signal Processing (ICASSP), Brighton, UK, May 12-17, 2019. [Demo, Code, Poster]