SQ-VAE: Variational Bayes on Discrete Representation with Self-annealed Stochastic QuantizationView Publication
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 PublicationsView All
STARSS23: An Audio-Visual Dataset of Spatial Recordings of Real Scenes with Spatiotemporal Annotations of Sound Events
Kazuki Shimada, Archontis Politis*, Parthasaarathy Sudarsanam*, Daniel Krause*, Kengo Uchida, Sharath Adavanne*, Aapo Hakala*, Yuichiro Koyama, Naoya Takahashi, Shusuke Takahashi, Tuomas Virtanen*, Yuki MitsufujiWhile direction of arrival (DOA) of sound events is generally estimated from multichannel audio data recorded […]
Automatic Piano Transcription with Hierarchical Frequency-Time Transformer
Keisuke Toyama, Taketo Akama, Yukara Ikemiya, Yuhta Takida, Wei-Hsiang Liao, Yuki MitsufujiTaking long-term spectral and temporal dependencies into account is essential for automatic piano transcriptio […]