Speech Synthesis and Recognition

Code for the project:

• Hi I am Ram (Text to Speech for different Dialects of Hindi)
• Dialect Recognition (Speech to Text for different Dialects of Hindi)

Overleaf link

Research Papers

• Deep Voice 3 - Oct 2017
• Wave Glow - Oct 2018
• Rule Based Grapheme to Phoneme Mapping for Hindi Speech Synthesis - 2003
• Hindi Datasets - 2017

Dataset

• Short Phrases with Audio
• Linguistic Data Consortium
• Corpus Linguistics
• Hindi Lessons

Code

• Rule based Grapheme to Phoneme model is ready
• Work in Progress for Segmentation Model (Encoder Decoder Architecture is being used with last layer in combination with GRU stacked attention)

To Read

• Creating a Dataset