spchlab is a Repository of Jupyter Noteboobs for Speech Recognition Demos, Tutorials and Exercises.
Different topics are organized by folder (LABxx_name). Each of these is rather independent. It is not strictly necessary to follow the order of the labs, but a higher numbered lab is likely to build on knowledge and skills acquired in lower numbered lab sessions.
Each lab-folder contains:
- a README.md file with a general description of the folder contents and links to the tutorials, demos and exercises
- a number of tutorial notebooks <Name>.ipynb. Web versions of these can be viewed directly via github-pages. The tutorial notebooks were also used for the generation of many figures in the course notes.
- a number of exercises in text <Ex_name>.md or in notebook <Ex_name>.ipynb format
For accessing the notebooks, follow the link to the correct LAB-folder.
- LAB00: Introduction
- LAB01: Speech and Audio
- LAB02: The Spectrogram
- LAB03: The Source Filter Model
- LAB04: Feature Extraction
- LAB05: Bayesian Classification
- LAB06: Frame Classification using GMMs and DNNs
- LAB07: Dynamic Time Warping
- LAB08: Hidden Markov Model Computations
- LAB09: Hidden Markov Model for Speech
The output of the tutorial notebooks may be found here