Jin Woo Lee

I am a Ph.D. candidate at Music and Audio Research Group of Seoul National University (SNU), advised by Professor Kyogu Lee. I work on audio signal processing and machine learning (ML). I did my undergrad in Mechanical Engineering at Pohang University of Science and Technology (POSTECH), where I spent most of my time studying computational fluid dynamics (CFD). Currently, I study characterizing and modeling the physical systems that give rise to the mechanism of acoustic propagation.

Education

Seoul National University (SNU)

Ph.D. in Intelligence and Information

Advisor: Kyogu Lee

2019-2024

Pohang University of Science and Technology (POSTECH)

B.Sc. in Mechanical Engineering

2013-2019

Experience

Meta | Reality Labs Research

Research Scientist Intern

Jun 2024 - Dec 2024 Redmond, WA

Supertone

Research Scientist Intern

2023, 2021 Seoul, KR

Research

My research interests include efficient simulation of wave propagation in the field of music and acoustics based on audio signal processing and ML. The research objectives can be broadly divided into two categories:

• Physics and ML-based acoustics modeling: includes musical acoustics simulation, room acoustics rendering and sound synthesis.

• Analysis on ML-based synthesis models: tackles characterizing synthesized data, such as perceptual evaluation metrics.

❦

2024년 1월 9일 11:49 오전

Inverse Nonlinearity Compensation of Hyperelastic Deformation in Dielectric Elastomer for Acoustic Actuation

Jin Woo Lee, Gwang Seok An, Jung-Yun Sun, Kyogu Lee, in manuscript under review [Paper]

We formulate a nonlinear ordinary differential equation governing the deformation of an elastomer actuator based on the hyperelastic model under ...

2023년 11월 23일 2:00 오후

String Sound Synthesizer on GPU-accelerated Finite Difference Scheme

Jin Woo Lee, Min Jun Choi, Kyogu Lee, in ICASSP 2024 [Paper] [Code] [Demo]

We open-source a PyTorch-based nonlinear string sound synthesizer, which operates on a finite difference time domain.

2023년 7월 15일 12:43 오전

AECSQI: Referenceless Acoustic Echo Cancellation Measures using Speech Quality and Intelligibility Improvement

Jin Woo Lee, Hyeong-Seok Choi, Kyogu Lee, in WASPAA 2023 (Oral presentation) [Paper] [Presentation]

We explore the utility of leveraging the speech quality improvement index for referenceless evaluation of acoustic echo cancellation models.

2023년 2월 16일 2:02 오후

Neural Fourier Shift for Binaural Speech Rendering

Jin Woo Lee, Kyogu Lee, in ICASSP 2023 [Paper] [Code] [Demo] [Poster]

We have proposed a fast and lightweight neural network for spatial audio rendering.

2023년 2월 16일 2:02 오후

Global HRTF Interpolation via Learned Affine Transformation of Hyper-conditioned Features

Jin Woo Lee, Sungho Lee, Kyogu Lee, in ICASSP 2023 [Paper] [Code] [Demo] [Poster]

We have explored a neural network-based interpolation method to address the sparsity issue in HRTF data.

2022년 3월 30일 12:02 오전

Representation Selective Self-distillation and wav2vec 2.0 Feature Exploration for Spoof-aware Speaker Verification

Jin Woo Lee, Eungbeom Kim, Junghyun Koo, and Kyogu Lee, in INTERSPEECH 2022 (Oral presentation) [Paper] [Presentation]

We explored what an effective feature space is for anti-spoofing and investigated methods to utilize it for secure ASV.

2018년 1월 18일 5:57 오후

Deep Learning Approach in Multi-scale Prediction of Turbulent Mixing-layer (Undergraduate Dissertation)

Jinu Lee, Sangseung Lee, and Donghyun You, in arXiv preprint | paper |

We tried to compress/learn the numerical simulation of turbulent fluid flow using lightweight neural networks.

❦

Academic Services

• Reviewer: Interspeech, ICASSP

• Teaching: Machine Listening (TA), Physical Modeling & Numerical Sound Synthesis (Mentor)

• Workshops: 2023 SANE Workshop (NYU), 2023 MA Workshop (SNU)

❦