I am doing a dissertation on Speech Enhancement using LSTM Recurrent neural netowrks. A large portion of this is background writing on the characteristics of speech and speech enhancement. I will take care of the the artificial intelligence bit, I just want the speech processing and other background bits taken care of.
below is a outline of what i need:
1. Charactistics of speech – how speech is created (doesnt need to be incredibly detailed) – charactistics of a speech signal – how we percieve speech
2. speech enhancement – cocktail party problem – speech enhancement algorithms (why some are good and bad)