Yohan Jayasanka

Yohan work on the area of speech recognition and intent identification. Currently, a sequential approach is used to identify the intent in speech. First, speech is converted into text using automatic speech recognizers ASRs, then this text is analyzed to identify the intent. In this cascaded approach, developing ASR is a huge bottleneck for low-resource languages. In this research, the possibility of direct speech intent identification is explored.