Time series generation (TSG) studies have mainly focused on the use of Generative Adversarial Networks (GANs) combined with recurrent neural network (RNN) variants. However, the fundamental limitations and challenges of training GANs still remain. In addition, the RNN-family typically has difficulties with temporal consistency between distant timesteps. Motivated by the successes in the image generation (IMG) domain, we propose TimeVQVAE, the first work, to our knowledge, that uses vector quantization (VQ) techniques to address the TSG problem. Moreover, the priors of the discrete latent spaces are learned with bidirectional transformer models that can better capture global temporal consistency. We also propose VQ modeling in a time-frequency domain, separated into low-frequency (LF) and high-frequency (HF). This allows us to retain important characteristics of the time series and, in turn, generate new synthetic signals that are of better quality, with sharper changes in modularity, than its competing TSG methods. Our experimental evaluation is conducted on all datasets from the UCR archive, using well-established metrics in the IMG literature, such as Fr\'echet inception distance and inception scores. Our implementation on GitHub: \url{https://github.com/ML4ITS/TimeVQVAE}.

翻译：时间序列生成（TSG）研究主要集中于生成对抗网络（GAN）与循环神经网络（RNN）变体的结合。然而，训练GAN的根本性限制与挑战仍然存在。此外，RNN系列通常难以处理远时间步之间的时序一致性。受图像生成（IMG）领域成功案例的启发，我们提出TimeVQVAE——据我们所知，这是首个利用矢量量化（VQ）技术解决TSG问题的研究。同时，离散潜空间的先验通过双向Transformer模型学习，能够更好地捕捉全局时序一致性。我们还在时频域中提出VQ建模，将其分离为低频（LF）和高频（HF）分量。这使得我们能够保留时间序列的重要特征，从而生成质量更高、模块化变化更锐利的新合成信号，优于其他TSG竞争方法。我们在UCR存档的所有数据集上进行了实验评估，采用图像生成领域公认的指标，如Fréchet初始距离（FID）和初始分数（IS）。我们的代码实现已托管于GitHub：\url{https://github.com/ML4ITS/TimeVQVAE}。