Preliminary study on using vector quantization latent spaces for TTS/VC systems with consistent performance
This incremental work addresses practical needs in speech synthesis for efficient data transfer and speaker anonymization.
The study explored using vector quantization for latent linguistic embeddings in TTS/VC systems, finding that it achieved similar performance in quality and speaker similarity with only a small degradation in perceptive evaluations, while enabling discrete latent spaces for reduced bit-rate and improved privacy.
Generally speaking, the main objective when training a neural speech synthesis system is to synthesize natural and expressive speech from the output layer of the neural network without much attention given to the hidden layers. However, by learning useful latent representation, the system can be used for many more practical scenarios. In this paper, we investigate the use of quantized vectors to model the latent linguistic embedding and compare it with the continuous counterpart. By enforcing different policies over the latent spaces in the training, we are able to obtain a latent linguistic embedding that takes on different properties while having a similar performance in terms of quality and speaker similarity. Our experiments show that the voice cloning system built with vector quantization has only a small degradation in terms of perceptive evaluations, but has a discrete latent space that is useful for reducing the representation bit-rate, which is desirable for data transferring, or limiting the information leaking, which is important for speaker anonymization and other tasks of that nature.