Speech recognition systems have been shown to be quite accurate in handling the way machines respond to human language. However, there are still many challenges that need to be overcome to improve the usefulness of these systems. These challenges arise from the different ways different people pronounce words. This might be as a result of different dialects being used, the person being a non-native speaker, or even just the variations between the spoken and written language. We can extract additional information from the human language with the help of some characteristic patterns in the speech signal known as acoustic cues. This project involves in describing a framework for labeling these acoustic cues in speech to enable a detailed analysis of cue modification patterns in speech.

“My personal interest relies on getting a better understanding of how speech recognition systems based on acoustic cues efficiently help improve the accuracy of the interaction between humans and machines. By the completion of this project, I also hope to have improved the programming and signal processing skills I started developing in Course 6.003 (Signals and Systems) I took in Spring 2019.”