Portrait of Anna (Cheng-Zhi) Huang

Anna (Cheng-Zhi) Huang

Affiliate Member
Canada CIFAR AI Chair
Assistant Professor, Massachussetts Institute of Technology (MIT), Department of Electrical Engineering and Computer Science
Université de Montréal
Research Scientist, Google Brain
Research Topics
Generative Models
Website
Google Scholar
X

Biography

Anna Huang is a Research Scientist at Google DeepMind, working on the Magenta project. She is also an Assistant Professor of Electrical Engineering and Computer Science at MIt's Faculty of Music and Theater Arts as well as an Adjunct Professor at Université de Montréal's Department of Computer Science and Operations Research (DIRO).

Her research focuses on designing generative models and interfaces to support music making and more generally the creative process. Her work is at the intersection of machine learning, human-computer interaction, and music. She is the creator of Music Transformer and Coconet. Coconet was the ML model that powered Google’s first AI Doodle, the Bach Doodle, which in two days harmonized 55 million melodies from users around the world. She was an organizer and judge for the international AI Song Contest for the past few years, and was a guest editor for TISMIR's special issue on AI and Musical Creativity. She holds a PhD from Harvard University, a master's from the MIT Media Lab, and a dual bachelor's degree in computer science and music composition from University of Southern California.

Current Students

Nithya Shikarpur
Master's Research - Université de Montréal
Co-supervisor :
Aaron Courville
snnithya@mit.edu
Website
Github
Google Scholar
Yusong Wu
PhD - Université de Montréal
Co-supervisor :
Aaron Courville

Publications

Adaptive Accompaniment with ReaLchords
Yusong Wu
Tim Cooijmans
Kyle Kastner
Adam Roberts
Ian Simon
Alexander Scarlatos
Chris Donahue
Cassie Tarakajian
Shayegan Omidshafiei
Aaron Courville
Pablo Samuel Castro
Natasha Jaques
Anna (Cheng-Zhi) Huang
Jamming requires coordination, anticipation, and collaborative creativity between musicians. Current generative models of music produce expr… (see more)essive output but are not able to generate in an online manner, meaning simultaneously with other musicians (human or otherwise). We propose ReaLchords, an online generative model for improvising chord accompaniment to user melody. We start with an online model pretrained by maximum likelihood, and use reinforcement learning to finetune the model for online use. The finetuning objective leverages both a novel reward model that provides feedback on both harmonic and temporal coherency between melody and chord, and a divergence term that implements a novel type of distillation from a teacher model that can see the future melody. Through quantitative experiments and listening tests, we demonstrate that the resulting model adapts well to unfamiliar input and produce fitting accompaniment. ReaLchords opens the door to live jamming, as well as simultaneous co-creation in other modalities.
2024-07-08
Proceedings of the 41st International Conference on Machine Learning (published)
proceedings.mlr.press
openreview.net
Grammar Generative Models for Music Notation
Anna (Cheng-Zhi) Huang
Deep generative models have been successfully applied in many learning experiments with digital data, such as images or audio. In the field … (see more)of music, they can also be used to generate symbolic representations, in the context of problems such as automatic music generation or transcription [1-3]. A significant challenge for generating structured symbolic data in general is obtaining well-formed results. This is especially true in the case of music. It is indeed widely accepted that musical notation represents, well beyond simple sequences of notes, a hierarchical organization of melodic and harmonic information, inducing non-local dependencies between musical objects [4]. A good representation of this information is essential for the interpretation and analysis of music pieces.
2023-01-01
(published)
www.semanticscholar.org
Improving Source Separation by Explicitly Modeling Dependencies between Sources
Ethan Manilow
Curtis Hawthorne
Anna (Cheng-Zhi) Huang
Bryan Pardo
Jesse Engel
We propose a new method for training a supervised source separation system that aims to learn the interdependent relationships between all c… (see more)ombinations of sources in a mixture. Rather than independently estimating each source from a mix, we reframe the source separation problem as an Orderless Neural Autoregressive Density Estimator (NADE), and estimate each source from both the mix and a random subset of the other sources. We adapt a standard source separation architecture, Demucs, with additional inputs for each individual source, in addition to the input mixture. We randomly mask these input sources during training so that the network learns the conditional dependencies between the sources. By pairing this training method with a blocked Gibbs sampling procedure at inference time, we demonstrate that the network can iteratively improve its separation performance by conditioning a source estimate on its earlier source estimates. Experiments on two source separation datasets show that training a Demucs model with an Orderless NADE approach and using Gibbs sampling (up to 512 steps) at inference time strongly outperforms a Demucs baseline that uses a standard regression loss and direct (one step) estimation of sources.
2022-05-23
ICASSP 2022 - 2022 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) (published)
doi.org
arxiv.org
Improving Source Separation by Explicitly Modeling Dependencies between Sources
Ethan Manilow
Curtis Hawthorne
Anna (Cheng-Zhi) Huang
Bryan A. Pardo
Jesse Engel
We propose a new method for training a supervised source separation system that aims to learn the interdependent relationships between all c… (see more)ombinations of sources in a mixture. Rather than independently estimating each source from a mix, we reframe the source separation problem as an Orderless Neural Autoregressive Density Estimator (NADE), and estimate each source from both the mix and a random subset of the other sources. We adapt a standard source separation architecture, Demucs, with additional inputs for each individual source, in addition to the input mixture. We randomly mask these input sources during training so that the network learns the conditional dependencies between the sources. By pairing this training method with a blocked Gibbs sampling procedure at inference time, we demonstrate that the network can iteratively improve its separation performance by conditioning a source estimate on its earlier source estimates. Experiments on two source separation datasets show that training a Demucs model with an Orderless NADE approach and using Gibbs sampling (up to 512 steps) at inference time strongly outperforms a Demucs baseline that uses a standard regression loss and direct (one step) estimation of sources.
2022-03-28
ArXiv (preprint)
doi.org
arxiv.org