Concatenative Speech Synthesis:

• Conversational Speech Synthesis, specially Disfluent Speech Synthesis
• Speech-to-Speech Translation, end-to-end systems, and how TTS systems can profit from source and how synthetic speech can be synchronised with source speech.
• Automatic Phonetic Speech Segmentation, mainly for the creation of concatenative TTS inventories, but also for automatic analysis of databases.
• Perceptual Evaluation of Synthetic Speech, evaluation of synthetic speech is an open problem still.

Natural Language Processing

Machine Learning

• Language and Speech processing are nowadays based on statistics and machine learning techniques such as:
• Artificial Neural Networks
• Classification and Regression Tree
• Language modelling