Our seminars take place on Wednesdays between 2:00 - 3.30 PM in the Center for Innovation and Transfer of Natural Sciences and Engineering Knowledge (building A0, wing B1, room 228).

Visit counter

Online: 1
Total: 61994

November 24, 2021, 2 p.m. room 228, dr inż. Leszek Gajecki, University of Information Technology and Management, Rzeszów, “Acoustic modelling for speech recognition”

LGajecki_300x200Abstract: The main aim of presented work is  acoustic modelling for telephone speech  recognition. The speech recognition of such speech as an example of conversational speech brings difficulty because of low quality of speech, different speakers voice, overlapping of speakers utterance, spontaneous talk with grammar errors. Our research was conducted using LUNA corpus consisting call centre dialogues in Polish language. The focus of research is the exploration of different models of neural networks and their combinations. The general architecture of speech recognition system will be presented as well as examples of known tools for speech recognition.


Comments are closed.