Sungho Lee

Ph.D. Candidate @ Music and Audio Research Group
Seoul National University
Email: sh-lee at snu.ac.kr
Scholar · GitHub · LinkedIn

Publications

Reverse Engineering of Music Mixing Graphs with Differentiable Processors and Iterative Pruning
Sungho Lee, Marco A. Martínez-Ramírez, Wei-Hsiang Liao, Stefan Uhlich, Giorgio Fabbro, Kyogu Lee, Yuki Mitsufuji
JAES, 2025
JAES · arXiv
Differentiable Artificial Reverberation
Sungho Lee, Hyeong-Seok Choi, Kyogu Lee
IEEE/ACM TASLP, 2022
demo · arXiv

Few-step Adversarial Schrödinger Bridge for Generative Speech Enhancement Seungu Han, Sungho Lee, Juheon Lee, Kyogu Lee
Interspeech, 2025 (Best student paper award nominee)
arXiv
TokenSynth: A Token-based Neural Synthesizer for Instrument Cloning and Text-to-Instrument
Kyungsu Kim, Junghyun Koo, Sungho Lee, Haesun Jeong, Kyogu Lee
ICASSP, 2025
demo · arXiv · code
Searching for Music Mixing Graphs: A Pruning Approach
Sungho Lee, Marco A. Martínez-Ramírez, Wei-Hsiang Liao, Stefan Uhlich, Giorgio Fabbro, Kyogu Lee, Yuki Mitsufuji
DAFx, 2024 (Best show & tell award)
demo · arXiv · code
Emosical: An Emotion-Annotated Musical Theatre Dataset
Hayoon Kim, Ahyeon Choi, Sungho Lee, Hyunjin Jung, Kyogu Lee
EMNLP Findings, 2024
ACL · code
Yet Another Generative Model for Room Impulse Response Estimation
Sungho Lee, Hyeong-Seok Choi, Kyogu Lee
WASPAA, 2023
demo · arXiv
Blind Estimation of Audio Processing Graph
Sungho Lee, Jaehyun Park, Seungryeol Paik, Kyogu Lee
ICASSP, 2023
demo · arXiv
Exploiting Time-Frequency Conformers for Music Audio Enhancement
Yunkee Chae, Junghyun Koo, Sungho Lee, Kyogu Lee
ACM MM, 2023
summary · arXiv · code
Global HRTF Interpolation via Learned Affine Transformation of Hyper-Conditioned Features
Jin Woo Lee, Sungho Lee, Kyogu Lee
ICASSP, 2023
summary · arXiv · code

GRAFX: An Open-Source Library for Audio Processing Graphs in PyTorch
Sungho Lee, Marco A. Martínez-Ramírez, Wei-Hsiang Liao, Stefan Uhlich, Giorgio Fabbro, Kyogu Lee, Yuki Mitsufuji
DAFx demo, 2024
arXiv · code · docs
Wavespace: A Highly Explorable Wavetable Generator
Hazounne Lee, Kihong Kim, Sungho Lee, Kyogu Lee
ISMIR LBD, 2024
arXiv · code (model) · code (VST)
Beat-Aligned Spectrogram-to-Sequence Generation of Rhythm-Game Charts
Jayeon Yi, Sungho Lee, Kyogu Lee
ISMIR LBD, 2023
demo · arXiv · code

Multi-modal Guitar Performance Transcription Apparatus and Method
Kyogu Lee, Sungho Lee, Chang-Bin Jeon, Jiwon Kim, Jongho Choi, Hayoon Kim, Yong-Ho Kim, Jang-Woo Lee
Domestic patent, 2025
DOI

Differentiable Acoustic Radiance Transfer
Sungho Lee, Matteo Scerbo, Seungu Han, Minjun Choi, Kyogu Lee, Enzo De Sena
arXiv
Rethinking Speech Representation Aggregation for Speech Enhancement with a Phonetic Mutual Information Perspective
Seungu Han, Sungho Lee, Kyogu Lee
UNISE: Unified Noise-Invariant Learning for Speech Enhancement toward Improved Content Preservation
Seungu Han, Sungho Lee, Eungbum Kim, Seaone Ok, Kyogu Lee
Solving Blind Non-linear Forward and Inverse Problems for Audio Applications
Jaehyun Park, Sungho Lee, Hayoon Kim, Kyogu Lee

Industry

Sony AI
Research intern, Summer 2023
Differentiable signal processing, music mixing, and graph learning.
Supertone
Research intern, Spring 2023
Impulse response estimation for audio post-production.
Product: Supertone Air

Funded projects

Invited talks

Music and AI Lab, National Taiwan University
Oct 7, 2024
Inverse Problems in Audio Post-production.
Center for Digital Music (C4DM), Queen Mary University of London
Nov 16, 2023
Deep Learning Meets Audio Processing Graphs.

Reviews