DataTalks.Club
DataTalks.Club
DataTalks.Club
Human-Centered AI for Disordered Speech Recognition - Katarzyna Foremniak
48 minutes Posted Oct 10, 2024 at 5:36 am.
DataTalks.Club intro
Background and career journey of Katarzyna
Transition from linguistics to computational linguistics
Merging linguistics and computer science
Understanding phonetics and morpho-syntax
Exploring morpho-syntax and its relation to grammar
Connection between phonetics and speech disorders
Improvement of voice recognition systems
Overview of speech recognition technology
Challenges of ASR systems with atypical speech
Strategies for improving recognition of disordered speech
Data augmentation for training models
Transfer learning in speech recognition
Challenges of collecting data for various speech disorders
Stammering and its connection to fluency issues
Polish consonant combinations and pronunciation challenges
Use of Amazon Transcribe for generating podcast transcripts
Role of language models in speech recognition
Contextual understanding in speech recognition
How voice recognition systems analyze utterances
Personalization of ASR models for individuals
Language disorders and their impact on communication
Applications of speech recognition technology
Challenges of personalized and universal models
Voice recognition in automotive applications
Humorous voice recognition failures in cars
Closing remarks and reflections on the discussion
0:00
48:01
Download MP3
Show notes
We talked about:
About the speaker:
Katarzyna is a computational linguist with over 10 years of experience in NLP and speech recognition. She has developed language models for automotive brands like Audi and Porsche and specializes in phonetics, morpho-syntax, and sentiment analysis.
Kasia also teaches at the University of Warsaw and is passionate about human-centered AI and multilingual NLP.
Join our slack: https://datatalks.club/slack.html