Authors

* External authors

Venue

Date

Share

SQ-VAE: Variational Bayes on Discrete Representation with Self-annealed Stochastic Quantization

Yuhta Takida

Takashi Shibuya

Wei-Hsiang Liao

Chieh-Hsin Lai

Junki Ohmura*

Toshimitsu Uesaka

Naoki Murata

Shusuke Takahashi*

Toshiyuki Kumakura*

Yuki Mitsufuji

* External authors

ICML 2022

2022

Abstract

One noted issue of vector-quantized variational autoencoder (VQ-VAE) is that the learned discrete representation uses only a fraction of the full capacity of the codebook, also known as codebook collapse. We hypothesize that the training scheme of VQ-VAE, which involves some carefully designed heuristics, underlies this issue. In this paper, we propose a new training scheme that extends the standard VAE via novel stochastic dequantization and quantization, called stochastically quantized variational autoencoder (SQ-VAE). In SQ-VAE, we observe a trend that the quantization is stochastic at the initial stage of the training but gradually converges toward a deterministic quantization, which we call self-annealing. Our experiments show that SQ-VAE improves codebook utilization without using common heuristics. Furthermore, we empirically show that SQ-VAE is superior to VAE and VQ-VAE in vision- and speech-related tasks.

Related Publications

Can Large Language Models Predict Audio Effects Parameters from Natural Language?

WASPAA, 2025
Seungheon Doh, Junghyun Koo*, Marco A. Martínez-Ramírez, Wei-Hsiang Liao, Juhan Nam, Yuki Mitsufuji

In music production, manipulating audio effects (Fx) parameters through natural language has the potential to reduce technical barriers for non-experts. We present LLM2Fx, a framework leveraging Large Language Models (LLMs) to predict Fx parameters directly from textual desc…

Large-Scale Training Data Attribution for Music Generative Models via Unlearning

ICML, 2025
Woosung Choi, Junghyun Koo*, Kin Wai Cheuk, Joan Serrà, Marco A. Martínez-Ramírez, Yukara Ikemiya, Naoki Murata, Yuhta Takida, Wei-Hsiang Liao, Yuki Mitsufuji

This paper explores the use of unlearning methods for training data attribution (TDA) in music generative models trained on large-scale datasets. TDA aims to identify which specific training data points contributed to the generation of a particular output from a specific mod…

Fx-Encoder++: Extracting Instrument-Wise Audio Effects Representations from Mixtures

ISMIR, 2025
Yen-Tung Yeh, Junghyun Koo*, Marco A. Martínez-Ramírez, Wei-Hsiang Liao, Yi-Hsuan Yang, Yuki Mitsufuji

General-purpose audio representations have proven effective across diverse music information retrieval applications, yet their utility in intelligent music production remains limited by insufficient understanding of audio effects (Fx). Although previous approaches have empha…

  • HOME
  • Publications
  • SQ-VAE: Variational Bayes on Discrete Representation with Self-annealed Stochastic Quantization

JOIN US

Shape the Future of AI with Sony AI

We want to hear from those of you who have a strong desire
to shape the future of AI.