Yu-Hua Chen

CS PhD candidate @ NTU | ML/Music/Guitar researcher | Prev: @PositiveGrid, @SonyAI_global

[Yu-Hua Chen @YuHuaChen12]

Excited to share two recent neural guitar amplifier modeling works, accepted at #DAFx2024 and #ISMIR2024 respectively:

1. Improving Unsupervised Clean-to-Rendered Guitar Tone Transformation Using GANs and Integrated Unaligned Clean Data
Paper Link: https://t.co/YxC33X180Y
Demo: https://t.co/Q8SVTSpC18
We investigate whether the performance of unsupervised guitar amplifier modeling can be improved by using different discriminators or integrating additional clean data. This work was done during an internship at @SonyAI_global .

2. Towards Zero-Shot Amplifier Modeling: One-to-Many Amplifier Modeling via Tone Embedding Control
Paper Link: https://t.co/EATjTpbMcy
Demo: https://t.co/EqtfdtHJic
We propose a one-to-many approach to model different sounds generated by various amplifiers using tone embedding conditioning control. We also provide a demo video demonstrating the flow of zero-shot tone cloning. This work was done during an internship at @PositiveGrid .

Looking forward to discussing all these with you at both conferences! Many thanks to the collaborators and the reviewers for their helpful feedback.

[Jack Loth @jackjamesloth]

Excited to announce that our paper "GOAT: A Large Dataset of Paired Guitar Audio Recordings and Tablatures" was accepted at @ISMIRConf this year in Korea!

paper: https://t.co/MDpM8xxD7x
dataset: https://t.co/9FwoAonDrp

[YenTung @YenTung11]

I’m pleased to share my internship work at @SonyAI_global, Fx-Encoder++, a representation for audio effects at mixture-level and instrument-wise level.
Arxiv: https://t.co/w3usSs8rZr
Code: https://t.co/DfWBHNqohH

[Yi-Hsuan Yang @affige_yang]

The revised version of METEOR, a new symbolic music re-orchestration model, has been accepted for publication at IJCAI'25 --- the AI, Arts and Creativity Special Track!
🧐Paper: https://t.co/9E00KGZpBt
🎻Demo: https://t.co/a3Yb9joLBc
🌟Code: https://t.co/kKPkjhTZVO

[Yu-Hua Chen @YuHuaChen12]

@PositiveGrid https://t.co/jTBScOdLPg

[arXiv Sound @ArxivSound]

``Towards Generalizability to Tone and Content Variations in the Transcription of Amplifier Rendered Electric Guitar Audio,'' Yu-Hua Chen, Yuan-Chiao Cheng, Yen-Tung Yeh, Jui-Te Wu, Jyh-Shing Roger Jang, Yi-Hsuan Yang, https://t.co/qRzoWbsm4U

[Yu-Hua Chen @YuHuaChen12]

🎸 New work on electric guitar transcription!

We introduce the EGDB-PG dataset (256 amp settings, 514 hours) with @PositiveGrid , and the Tone-informed Transformer model using tone embeddings.

Check out our in-the-wild demo!

Paper: https://t.co/HhIopu9MEx
Demo & dataset: https://t.co/ezS7Cuwi9Z

[Bo-Yu Chen @Chen_Paul_u]

🚀 Introducing AI TrackMate at @NeurIPSConf 2024 Creative AI Track! 🚀

🎧 Drop in your audio, AI TrackMate listen, chat, and guide you with insightful feedback and tips to transform your tracks!

🌐 Demo: https://t.co/Q5yebGUQnC
📄 Paper: https://t.co/h6C5me2wxt

#NeurIPS2024

[arXiv Sound @ArxivSound]

``Open-Amp: Synthetic Data Framework for Audio Effect Foundation Models,'' Alec Wright, Alistair Carson, Lauri Juvela, https://t.co/kVqrFWTmkJ

[Yu-Hua Chen @YuHuaChen12]

We just set up everything for showcasing our LBD paper. If you are interested in guitar tone modeling, come join us for a discussion.
#ismir2024

[Pedro Sarmento @umpedronosapato]

We bring progressive metal madness to @ISMIRConf 🤘 really glad 'Between the AI and Me' was accepted and we were able to present our work to this titanic community

@btbamofficial

[Pedro Sarmento @umpedronosapato]

Insanely good work by @YuHuaChen12 @affige_yang @PositiveGrid at @ISMIRConf 🤘🎸 awesome choice of sound examples too!

[Yu-Hua Chen @YuHuaChen12]

@umpedronosapato @affige_yang @PositiveGrid @ISMIRConf Thank you my man!

[arXiv Sound @ArxivSound]

``Music Foundation Model as Generic Booster for Music Downstream Tasks,'' WeiHsiang Liao, Yuhta Takida, Yukara Ikemiya, Zhi Zhong, Chieh-Hsin Lai, Giorgio Fabbro, Kazuki Shimada, Keisuke Toyama, Kinwai Cheuk, Marco A. Mart\'inez-Ram\'irez, Shusuke Takaha… https://t.co/yrGBaBkTgQ

[Yu-Hua Chen @YuHuaChen12]

We are happy to announce that the short version of our zero-shot amplifier modeling paper has been accepted to the #ismir2024 LBD session. We plan to showcase the functionality of the plugin version of our neural network model during the session (and yes, we’ll bring a guitar and speaker!). Please visit us at the LBD session!

Shout out to @PeripheryBand for the great song!

Paper Link: https://t.co/FOcX7BqR3M

[シュナプーン @schnapoon]

パラパラAI漫画を作りたい人はStable Diffusionがオススメ。これ実は個別のAI生成画像を連続させただけのもの。こんな風に違和感なく一貫性の高い画像を連続的に出力できる。誰もがPCで無償で活用できるのが魅力的。 作り方をリプ欄にまとめたので、少しだけでも見て。

[Marco Martínez @marcoamaram]

And happy for the Best Show & Tell Award ! It feels good that a mixing paper got some love at DAFx😊

Congrats @sh_lee_97 and all our collaborators at Sony AI and SNU ! 🎉

Check out our demo and open-source library—GRAFX

https://t.co/pkHYSvUk0y

https://t.co/XcSTinaLd0

[Marco Comunità @marcomunita]

AFX-Research: an Extensive and Flexible Repository of Research about Audio Effects

All the research on audio effects of the last few decades. A table with lots of metadata. Search. Filter. Order. Contribute.

repo: https://t.co/NPGj6g28on
web: https://t.co/sU92smDy45

[Jesse Engel @jesseengel]

As a guitarist myself, it's really cool to see the DDSP approach being valuable all the way down to modeling the interiors of a guitar amp 🤘

[Christian Steinmetz @csteinmetz1]

DDSP Guitar Amp: Interpretable Guitar Amplifier Modeling

https://t.co/9usnUh8Lky

[Jingyue Huang @Yuer867]

Excited to share our new emotion-driven music generation model 🎶EMO-Disentanger🎶! Published at #ISMIR2024 🥳

📍Emotion-driven Piano Music Generation via Two-stage Disentanglement and Functional Representation
📑Paper: https://t.co/Yr1xfyqnmE
🎵Demo: https://t.co/FpRREJQNAM

[YenTung @YenTung11]

Audio effect modeling with knob tweaking is key to match how musicians & engineers actually use audio effects in their workflow. But how do we feed this info into the model? 🤔
Check out our paper: accepted at DAFx24! 🎉
https://t.co/74hDEO5dHH