Editing Advancements in AI TTS (2020s) (section)

=== 2. Efficient Speech Synthesis<ref name=":0" /> ===
It is about how to reduce the cost of speech synthesis including the cost of collecting and labeling training data, training and serving TTS models, etc.

==== 2.1 Data-efficient TTS ====
Many '''low-resource languages''' are lack of training data. How to leverage unsupervised/semi-supervised learning and cross-lingual transfer learning to help the low-resource languages is an interesting direction. For example, the [https://www.zerospeech.com/ ZeroSpeech Challenge] is a good initiative to explore the techniques to learn only from speech, without any text or linguistic knowledge. Besides, in voice adaptation, a target speaker usually has little adaptation data, which is another application scenario for data-efficient TTS. 

==== 2.2 Parameter-efficient TTS ====
Today’s neural TTS systems usually employ large neural networks with tens of millions of parameters to synthesize high-quality speech, which block the applications in mobile, low-end devices due to their limited memory and power consumption. Designing compact and lightweight models with less memory footprints, power consumption and latency are critical for those application scenarios. 

==== 2.3 Energy-efficient TTS ====
Training and serving a high-quality TTS model consume a lot of energy and emit a lot of carbon. Improving energy efficiency, e.g., reducing the [[wikipedia:FLOPS|FLOPs]] in TTS training and inference, is important to let more populations to benefit from advanced TTS techniques while reducing carbon emissions to protect our environment.