Overview
Explore the fundamentals of speech recognition systems in this PyCon US talk by Javier Jorge Cano. Delve into the technical challenges and processes behind converting voice signals to text, including feature extraction, acoustic and language modeling, and recognition graph creation. Learn how to develop a system for recognizing isolated words in Python through a guided example. Gain insights into the implementation of advanced speech recognition techniques in production systems, providing transcriptions for various companies and institutions. Discover open resources, toolkits, and evaluation methods used in Automatic Speech Recognition (ASR). Perfect for developers interested in understanding and implementing speech recognition technology using Python.
Syllabus
Intro
What is this talk about?
Some remarks...
Automatic Speech Recognition (ASR)
ASR is not already solved?
Technical Challenges
The big picture
ASR System: Steps
ASR System: Toolkits
ASR System: Open Resources
ASR System: Prepare resources
ASR System: Resources
ASR System: Feature Extraction
ASR System: Modelling
ASR System: Acoustic Modelling
ASR System: Language Modelling
ASR System: Recognition graph
ASR System: Evaluation
ASR System: Advanced Topics
What was this talk about?
Taught by
PyCon US