Reo YONEYAMA

Photograph of me
Ph.D. in Informatics
Research Engineer, AI Lab, CyberAgent, Inc.
yoneyama_reo [at] cyberagent.co.jp

EDUCATION

Apr. 2021 - Mar. 2026
TMI WISE Program: Graduate Program for Lifestyle Revolution Based on Transdisciplinary Mobility Innovation
Apr. 2023 - Mar. 2026
Ph.D., Department of Intelligent Systems, Graduate School of Informatics, Nagoya University, Japan (Supervisor: Prof. Tomoki Toda)
Apr. 2021 - Mar. 2023
M.S., Department of Intelligent Systems, Graduate School of Informatics, Nagoya University, Japan
Apr. 2017 - Mar. 2021
B.S., Department of Computer Science, School of Informatics, Nagoya University, Japan
Mar. 2017
Graduated from Tochigi Prefectural Kanuma High School, Japan

WORK EXPERIENCE

Apr. 2026 - present
Research Engineer, Audio Nexus Team, AI Lab, CyberAgent, Inc., Japan
May 2025 - Mar. 2026
R&D Internship, Speech & Language Team, SB Intuitions Corp., Japan
July 2024 - Mar. 2025
Joint Research Project with the ASP team, LY Corporation, Japan
May 2022 - Oct. 2025
R&D Internship, TARVO Inc., Japan
Feb. 2023 - Mar. 2024
Joint Research Project with the ASP team, LY Corporation, Japan
May 2021 - Jan. 2023
Research Internship, Voice Team, LINE Corporation, Japan
Dec. 2019 - Apr. 2021
R&D Internship, Acompany Co., Ltd., Japan

JOURNALS

IEICE Trans. Inf. Syst.
SiFi-GAN: Combining Source-Filter Modeling and Upsampling-Based High-Fidelity Neural Vocoder for Fast and Pitch-Controllable Speech Synthesis [link]

Reo Yoneyama, Tomoki Toda

IEEE TASLP
Wavehax: Aliasing-Free Neural Waveform Synthesis Based on 2D Convolution and Harmonic Prior for Reliable Complex Spectrogram Estimation [link]

Reo Yoneyama, Atsushi Miyashita, Ryuichi Yamamoto, Tomoki Toda

IEEE/ACM TASLP
High-Fidelity and Pitch-Controllable Neural Vocoder Based on Unified Source-Filter Networks [link]

Reo Yoneyama, Yi-Chiao Wu, Tomoki Toda

CONFERENCE PROCEEDINGS (PEER-REVIEWED)

EUSIPCO 2025
VAE-SiFiGAN: Source-Filter HiFi-GAN Based on Variational Autoencoder Representations with Enhanced Pitch Controllability

Kenichi Ogita, Reo Yoneyama, Wen-Chin Huang, Tomoki Toda

Interspeech 2025
Comparative Analysis of Fast and High-Fidelity Neural Vocoders for Low-Latency Streaming Synthesis in Resource-Constrained Environments

Reo Yoneyama, Masaya Kawamura, Ryo Terashima, Ryuichi Yamamoto, Tomoki Toda

ASRU 2023
A Comparative Study of Voice Conversion Models with Large-Scale Speech and Singing Data: The T13 Systems for the Singing Voice Conversion Challenge 2023 [link]

Ryuichi Yamamoto, Reo Yoneyama, Lester Phillip Violeta, Wen-Chin Huang, Tomoki Toda

IEEE ICASSP 2023
Source-Filter HiFi-GAN: Fast and Pitch Controllable High-Fidelity Neural Vocoder [link]

Reo Yoneyama, Yi-Chiao Wu, Tomoki Toda

IEEE ICASSP 2023
Nonparallel High-Quality Audio Super Resolution with Domain Adaptation and Resampling CycleGANs [link]

Reo Yoneyama, Ryuichi Yamamoto, Kentaro Tachibana

IEEE ICASSP 2023
NNSVS: A Neural Network-Based Singing Voice Synthesis Toolkit [link]

Ryuichi Yamamoto, Reo Yoneyama, Tomoki Toda

Interspeech 2022
Unified Source-Filter GAN with Harmonic-plus-Noise Source Excitation Generation [link]

Reo Yoneyama, Yi-Chiao Wu, Tomoki Toda

Interspeech 2021
Unified Source-Filter GAN: Unified Source-Filter Network Based On Factorization of Quasi-Periodic Parallel WaveGAN [link]

Reo Yoneyama, Yi-Chiao Wu, Tomoki Toda

CONFERENCE PROCEEDINGS (NON PEER-REVIEWED)

ASA/ASJ 2025
Why is a sinusoidal signal input effective in time-domain neural vocoders?

Reo Yoneyama, Tomoki Toda

ASJ Spring 2025
Aliasing-Free Neural Vocoder Based on Complex Spectrogram Estimation with Harmonic Signal Modeling and 2D Convolution

Reo Yoneyama, Atsushi Miyashita, Ryuichi Yamamoto, Tomoki Toda

ASJ Spring 2025
VAE-SiFi-GAN: SiFi-GAN Based on Variational Autoencoder Representations

Kenichi Ogita, Reo Yoneyama, Wen-Chin Huang, Tomoki Toda

ASJ Autumn 2023
NNSVS: A Neural Network-Based Singing Voice Synthesis Toolkit [link]

Ryuichi Yamamoto, Reo Yoneyama, Tomoki Toda

ASJ Spring 2023
Source-Filter-Architecture-Based HiFi-GAN

Reo Yoneyama, Yi-Chiao Wu, Tomoki Toda

ASJ Spring 2022
Improvement of Unified Source-Filter Network with Adversarial Learning

Reo Yoneyama, Yi-Chiao Wu, Tomoki Toda

ASJ Autumn 2021
Unified Source-Filter Network with Adversarial Learning

Reo Yoneyama, Yi-Chiao Wu, Tomoki Toda

IEICE Tech. Rep. 2021
A unified source-filter network for neural vocoder [paper]

Reo Yoneyama, Yi-Chiao Wu, Tomoki Toda

PREPRINTS

arXiv 2026
Sarashina2.2-TTS: Tackling Kanji Polyphony in Japanese Speech Generation via Data Scaling and Targeted Data Synthesis [link]

Lianbo Liu, Shiao Zhu, Kai Washizaki, Reo Yoneyama, Haesung Jeon, Mengjie Zhao, Yusuke Fujita, Hao Shi, Nao Yoshida, Yuan Gao, Roman Koshkin, Yukiya Hono, Yui Sudo

INVITED TALKS

June 2025
Invited talk at the 143rd MUS & 156th SLP Joint Workshop [slides]
“Overview of Neural Vocoders: From the Perspective of Generative Models and Practicality”

HONORS & AWARDS

Mar. 2026
Excellent Doctor Award, Graduate School of Informatics, Nagoya University, Japan
Jan. 2024
IEEE SPS Japan Student Conference Paper Award 2023 [link]
Oct. 2022
2022 Tokai Area Speech-Related Research Laboratory Master's Midterm Presentation Q&A Award
Sep. 2021
23rd Japan Acoustics Society Student Excellent Presentation Award [link]
Nov. 2019
Domestic Qualifier of International Collegiate Programming Contest (ICPC) Asia Yokohama Regional 2019 [link]

GRANTS

Apr. 2024 - Mar. 2026
Japan Society for the Promotion of Science (JSPS) Research Fellowship for Young Scientists (DC2) [link]
Apr. 2023 - Mar. 2024
Nagoya University Interdisciplinary Frontier Fellowship (declined remainder to accept JSPS DC2) [link]
Apr. 2021 - Mar. 2023
TMI WISE Program Fellowship (declined remainder to accept the Nagoya University Interdisciplinary Frontier Fellowship) [link]

LANGUAGES

Japanese
Native
English
Intermediate (TOEIC Listening & Reading Test, Taken in May 2022)
Score: 865 (Listening: 430, Reading: 435)

MISCELLANEOUS

Technical Advising at DubGuild Inc. (Apr. 2025 – Mar. 2026) and for Melisma Software, Developed by Sakana Nakasako (Dec. 2023 – Oct. 2025; paper). Competitive Programming: AtCoder (Highest rating: 1545), ICPC Asia Yokohama Regional (2019) and Domestic Qualifier (2020), and JPHACKS @ Nagoya (2019).
expand_less